JP2017122462A - Driving device for fuel injection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a fuel injection device driving device enabled to enhance the precision of a fuel injection quantity by combining fuel injections with a plurality of injection pulse widths.SOLUTION: In a fuel injection device driving device having a function to drive a fuel injection device so that a fuel injection in one combustion cycle may be divided into a plurality of times, the fuel injection device is driven such that a plurality of split injections to be performed in one combustion cycle may contain a fuel injection in a target opening for a valve body or a movable element of the fuel injection device to reach a restricting part or a fuel injection for the valve body to fail in reaching the restricting part.SELECTED DRAWING: Figure 36

Description

  The present invention relates to a drive device for driving a fuel injection device of an internal combustion engine.

  As a background art in this technical field, there is JP 2013-144971 A (Patent Document 1). This publication provides a control device for an internal combustion engine capable of highly accurate fuel injection control when a fuel injection valve (fuel injection device) is half lifted (a state in which the fuel injection device is closed before it is fully opened). For the purpose, the following configuration is described. A control device for an internal combustion engine described in Patent Document 1, a fuel injection valve that supplies fuel to the internal combustion engine, a means for calculating an energization time of the fuel injection valve, a sensor that detects opening of the fuel injection valve, A means for calculating and storing a valve opening delay time, which is a difference between an energization start time for the fuel injection valve and a valve opening detection time, and when the energization time for the fuel injection valve is greater than or equal to a predetermined value by the energization time determination unit The valve opening delay time is calculated and stored, and when the energization time to the fuel injection valve is less than a predetermined value, the energization time of the fuel injection valve is controlled based on the valve opening delay time stored in the storage means, Control to increase the energization time of the fuel injection valve (see summary). Note that Patent Document 1 considers that the amount of fuel required per cycle is divided into a plurality of injections (see paragraph 0023).

JP2013-144971A

  It is known that a fuel injection valve can perform highly accurate fuel injection by using a region (linear region) in which an energization time (injection pulse width) of a drive signal and a fuel injection amount change linearly. On the other hand, the fuel injection in the half lift state exists in a region (non-linear region) where the amount of fuel injected does not change linearly with respect to the change in the injection pulse width.

  In Patent Document 1, when the amount of fuel required per cycle is divided and injected several times, the fuel injection tank valve injects fuel in a half lift state, and energizes during half lift. A technique for controlling the amount of injected fuel with high accuracy by correcting the time is disclosed. However, in Patent Document 1, there is no consideration for increasing the accuracy of the injected fuel injection amount by combining the fuel injection with a plurality of injection pulse widths when the fuel amount is divided and injected multiple times. .

  The objective of this invention is providing the drive device of the fuel-injection apparatus which can improve the precision of the fuel-injection amount injected by combining the fuel injection by a several injection pulse width.

  In order to solve the above-described problem, the fuel injection device drive device according to the present invention includes a fuel injection by a full lift in a plurality of divided injections when the fuel injection performed in one combustion cycle is divided into a plurality of times. Fuel injection is performed so as to include fuel injection by half lift. Here, the full lift is a state in which the valve body is displaced (opened) to a state where the displacement is regulated by the regulating portion that regulates the displacement amount of the valve body, and is opened when the valve body reaches the regulating portion. Refers to the valve state. Moreover, a half lift means the valve opening state by the intermediate opening degree which a valve body does not reach | attain a control part.

  According to the present invention, a plurality of divided injections performed in one combustion cycle include a fuel injection by a half lift with a low control accuracy of the injection amount and a fuel injection by a full lift capable of controlling the injection amount with high accuracy. Thus, the total amount of fuel injected by the plurality of divided injections can be brought close to the target injection amount. Thereby, the drive device of the fuel injection device that can improve the accuracy of the injected fuel injection amount can be provided.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a longitudinal sectional view of a fuel injection device according to a first embodiment of the present invention and a configuration of a drive circuit and an engine control unit (ECU) connected to the fuel injection device. It is the figure which showed the cross-sectional enlarged view of the drive part structure of the fuel-injection apparatus in 1st Example. It is the figure which showed the relationship between the injection pulse which drives the fuel-injection apparatus in 1st Example, the voltage between terminals applied to the solenoid of a fuel-injection apparatus, a drive current, a valve body, and a needle | mover displacement amount, and time. FIG. 4 is a diagram showing a relationship between an injection pulse width Ti output from an ECU in FIG. 3 and a fuel injection amount injected from a fuel injection device. It is the figure which showed the relationship between the injection pulse width Ti and fuel injection quantity of a fuel injection apparatus with individual variation in injection quantity characteristics. It is the figure which showed the valve behavior in each point 501, 502, 503, 531, 532 in FIG. It is the figure which showed the relationship between the injection pulse width Ti output from a drive device, a drive current, the displacement amount of a valve body, the amount of displacement of a needle | mover, and time. It is the figure which showed the detail of the drive device and ECU (engine control unit) of a fuel-injection apparatus. Injection pulse width Ti, driving current, current differential value, current second-order differential value, valve body displacement amount, mover of three fuel injection devices having different valve timings due to the influence of dimensional tolerance variations in the first embodiment It is the figure which showed the relationship between displacement amount and time. In the first embodiment, the injection pulse Ti, the drive current supplied to the fuel injection device, the operation timing of the switching element of the drive device, the voltage Vinj between the solenoid terminals, the displacement amount of the valve body and the mover, the relationship between the mover acceleration and time FIG. The driving current supplied to the solenoid 105 in the first embodiment, the displacement amounts of the three individual valve bodies whose valve closing behavior differs depending on the dimensional tolerance of the fuel injection device, the enlarged view of the voltage VL1, and the second-order differential value of the voltage VL1 FIG. The figure which showed the correspondence of the magnetic flux (phi) passing through the attraction | suction surface between the needle | mover and the fixed core of a needle | mover, and the voltage Vinj between solenoid terminals between the needle | mover and fixed core in 1st Example. It is. Under the condition that the valve body in the first embodiment reaches the target lift, the voltage Vinj between the terminals, the drive current, the first-order differential value of the current, the current It is the figure which showed the relationship between a 2nd-order differential value, a valve body displacement amount, and time. It is the figure which showed the initial stage magnetization curve and return curve of the magnetization curve (BH curve) of the magnetic material used for a magnetic circuit in a 1st Example. It is the figure which described the flowchart of the injection amount correction method of each cylinder in the area | region where the injection pulse width Ti used as the intermediate | middle lift area where the valve body in 1st Example does not reach | attain a target lift is small. When the injection pulse width Ti is changed under certain fuel pressure conditions in the first embodiment, the injection amount of each cylinder, the valve closing completion timing Tb, the valve opening start timing Ta ′, and the unit time injected from the fuel injection device It is the figure which showed the relationship of the detection information (Tb-Ta ') * Qst calculated | required from flow volume Qst (henceforth a static flow). It is the figure which showed the relationship between the detection information of the fuel injection device of each cylinder in the 1st example, individual 2 and solid 3, and injection pulse width Ti. Injection pulse width Ti, drive current, inter-terminal voltage ∨inj, second-order differential value of voltage VL1, and second-order differential value of current, that is, voltage VL2, under the condition for dividing injection performed during one intake / exhaust stroke in the first embodiment It is the figure which showed the amount of displacement of a valve body, and the relationship of time. It is an enlarged view of the section of a drive part in the valve closing state where the valve element of the fuel injection device in the 2nd example is contacting the valve seat. It is the figure which expanded the longitudinal cross-section of the valve body front-end | tip part of the fuel-injection apparatus in a 2nd Example. It is an enlarged view of the section of a drive part in the valve open state of the valve element of the fuel injection device in the second example of the present invention. It is an enlarged view of a section of a drive part at the moment when a valve body of a fuel injection device in a second example starts valve closing from a valve-opened state and contacts valve seat 118. It is the figure which showed the structure of the drive device in 2nd Example. This is the potential difference between the terminal for detecting the displacement and voltage VL of the second valve body and the second mover when the valve is closed from the maximum lift in the intermediate lift state in the second embodiment and the ground potential. It is the figure which showed the relationship between the voltage VL4, the 2nd-order differential value of voltage VL4, and the time after injection pulse OFF. Among the cases where the fuel injection device is driven by the method of the third embodiment, the voltage Vinj between the terminals of the fuel injection device and the drive current when the valve body or the second valve body is used while being held at the target lift position for a certain period of time. , Magnetic attraction force acting on the mover or the second mover, valve element driving force acting on the valve element or the second valve element, displacement amount of the valve element or the second valve element, mover or second It is the figure which showed the relationship between the displacement amount of a needle | mover, and time. Among the cases where the fuel injection device is driven by the method of the third embodiment, the inter-terminal voltage Vinj in the operating state when the minimum injection amount is performed while the valve body or the second valve body reaches the target lift. , Driving current, magnetic attractive force acting on the valve mover or the second mover, valve drive force acting on the valve element or the second valve element, displacement amount of the valve element or the second valve element, mover Or it is the figure which showed the relationship between the displacement amount of the 2nd needle | mover, and time. Among the cases where the fuel injection device is driven by the method of the third embodiment, the voltage Vinj between the terminals when operating with the intermediate lift, the drive current, the magnetic attractive force acting on the mover or the second mover, and the valve body Or it is the figure which showed the relationship between the valve body drive force which acts on a 2nd valve body, the displacement amount of a valve body or a 2nd valve body, the displacement amount of a needle | mover or a 2nd needle | mover, and time. Further, in the drawing of the valve body driving force, the driving force in the valve opening direction is shown in the positive direction, and the driving force in the valve closing direction is shown in the negative direction. It is the figure which showed the relationship between the injection pulse width Ti and the fuel injection quantity q at the time of using the current waveform of the control system of FIGS. 27-29 of 3rd Example. The injection period (Tb−Ta ′) should be the same for individuals having different valve opening start timing Ta ′ and valve closing completion timing Tb of the valve body or the second valve body under the condition of supplying the same injection pulse width Ti. FIG. 5 is a diagram showing the relationship between the drive voltage, drive current, valve body displacement amount and time of each individual as a result of correcting the injection pulse, drive voltage, and drive current. It is the figure which described the adjustment method of the injection amount after adjusting the injection period in the minimum injection amount in 4th Example. It is the figure which showed the relationship between the injection pulse after adjusting the injection period in the minimum injection quantity in 4th Example, and injection quantity. It is a block diagram of the cylinder direct injection type gasoline engine in 5th Example. It is a figure which shows the structure of the longitudinal cross-sectional view of the fuel-injection apparatus of 6th Example. The voltage between the terminals of the solenoid when using the fuel injection device of the sixth embodiment, the drive current supplied to the solenoid, the difference between the current value under the condition that the valve element does not open and the current value of each individual, and the valve displacement It is the figure which showed the relationship between the time after injection pulse ON. It is explanatory drawing of the fuel injection timing correction method. The relationship among the injection pulse, the injection period, and the valve lift is shown for the case where the fuel injection in one combustion cycle is divided into a plurality of times. It shows the change in engine speed from cranking to fast idle section. It shows changes in the engine speed and the fuel pressure supplied to the fuel injection device from the cranking to the fast idle section in the ninth embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  Hereinafter, the operation of the fuel injection system including the fuel injection device and the drive device according to the present invention will be described with reference to FIGS.

  First, the configuration and basic operation of the fuel injection device and its driving device will be described with reference to FIG. FIG. 1 is a longitudinal sectional view of a fuel injection device, and a diagram showing an example of the configuration of a drive circuit 121 and an ECU (engine control unit) 120 for driving the fuel injection device. In this embodiment, the ECU 120 and the drive circuit 121 are configured as separate devices, but the ECU 120 and the drive circuit 121 may be configured as an integrated device. A device constituted by the ECU 120 and the drive circuit 121 will be described below as a drive device.

  The ECU 120 takes in signals indicating the state of the engine from various sensors, and calculates the injection pulse width and the injection timing for controlling the injection amount injected from the fuel injection device in accordance with the operating conditions of the internal combustion engine. The injection pulse output from the ECU 120 is input to the drive circuit 121 of the fuel injection device through the signal line 123. The drive circuit 121 controls the voltage applied to the solenoid 105 and supplies a current. The ECU 120 communicates with the drive circuit 121 through the communication line 122 to switch the drive current generated by the drive circuit 121 according to the pressure of the fuel supplied to the fuel injection device and the operation conditions, and to set current and time values. It is possible to change. The drive circuit 121 can change the control constant by communication with the ECU 120, and can change the set value of the current waveform according to the control constant.

  Next, the configuration and operation of the fuel injection device will be described with reference to a vertical cross section of the fuel injection device in FIG. 1 and an enlarged cross-sectional view of the vicinity of the movers 102a and 102b and the movable member 114 in FIG. Note that. The mover 102a and the mover 102b may be configured as an integral part. A component composed of the movable element 102a and the movable element 102b is referred to as a movable element 102. The fuel injection device shown in FIGS. 1 and 2 is a normally closed electromagnetic valve (electromagnetic fuel injection device), and when the solenoid (coil) 105 is not energized, the spring 110 is a first spring. Accordingly, the movable element 102b is urged in the valve closing direction, and the end surface 207 of the movable element 102b on the valve body 114 side and the upper end surface of the valve body 114 are in contact with each other. At this time, since the load due to the set spring 110 acts on the valve body 114 via the movable element 102b, the valve body 114 is urged toward the valve seat 118 and closes in close contact with the valve seat 118. It is in a state. In the valve-closed state, the force by the spring 110 applied in the valve closing direction and the force by the return spring 112 of the second spring applied in the valve opening direction act on the movable element 102. At this time, since the force by the spring 110 is larger than the force by the return spring 112, the end surface 207 of the movable element 102b contacts the valve body 114, and the movable element 102 is stationary. In the valve-closed state, there is a gap 201 between the contact surface 205 of the valve body 114 with the movable element 102a and the movable element 102a. In this state, there is a gap between the mover 102 and the fixed core 107. Further, the valve body 114 and the mover 102 are configured to be relatively displaceable and are contained in the nozzle holder 101. In addition, the nozzle holder 101 has an end surface 208 that serves as a spring seat for the return spring 112. The force by the spring 110 is adjusted at the time of assembly by the pushing amount of the spring retainer 124 fixed to the inner diameter of the fixed core 107. The urging force of the zero position spring 112 is set smaller than the urging force of the spring 110.

  In the fuel injection device, the fixed core 107, the mover 102, the nozzle holder 101, and the housing 103 constitute a magnetic circuit, and a gap is provided between the mover 102 and the fixed core 107. A magnetic diaphragm 111 is formed in a portion corresponding to the gap between the mover 102 and the fixed core 107 of the nozzle holder 101. The solenoid 105 is attached to the outer peripheral side of the nozzle holder 101 while being wound around the bobbin 104. A rod guide 115 is provided in the vicinity of the tip of the valve body 114 on the valve seat 118 side so as to be fixed to the nozzle holder 101. The rod guide 115 may be configured as the same part as the orifice cup 116. The valve body 114 is guided in movement in the valve axis direction by two rod guides, a first rod guide 113 and a second rod guide 115. An orifice cup 116 in which a valve seat 118 and a fuel injection hole 119 are formed is fixed at the tip of the nozzle holder 101, and an internal space (fuel passage) in which the movable element 102 and the valve body 114 are provided from the outside. It is sealed.

  The fuel supplied to the fuel injection device is supplied from a rail pipe provided upstream of the fuel injection device, flows to the tip of the valve body 114 through the first fuel passage hole 131, and the valve seat 118 of the valve body 114. The fuel is sealed by the seat portion formed at the side end portion and the valve seat 118. When the valve is closed, the pressure difference between the upper part and the lower part of the valve body 114 is generated by the fuel pressure, and the valve body 114 is pushed in the valve closing direction by the force multiplied by the fuel pressure and the pressure receiving surface of the seat inner diameter at the valve seat position. In the closed state, a gap 201 is provided between the contact surface 205 of the valve element 114 with the movable element 102a and the movable element 102a. When a current is supplied to the solenoid 105, a magnetic flux passes between the fixed core 107 and the movable element 102 by a magnetic field generated by the magnetic circuit, and a magnetic attractive force acts on the movable element 102. At the timing when the magnetic attractive force acting on the mover 102 exceeds the load by the set spring 110, the mover 102 starts to move in the direction of the fixed core 107. At this time, since the valve body 114 and the valve seat 118 are in contact with each other, the movement of the mover 102 is performed in a state where there is no fuel flow, and is separated from the valve body 114 receiving the differential pressure due to the fuel pressure. Therefore, it is possible to move at high speed without being affected by fuel pressure or the like.

  When the amount of displacement of the movable element 102 reaches the size of the gap 201, the movable element 102 transmits a force to the valve body 114 through the contact surface 205, and lifts the valve body 114 in the valve opening direction. At this time, since the movable element 102 performs idle running and collides with the valve body 114 in a state having kinetic energy, the valve body 114 receives the kinetic energy of the movable element 102 and rapidly moves in the valve opening direction. Start displacement. A differential pressure generated with the fuel pressure acts on the valve body 114, and the differential pressure acting on the valve body 114 is within a range where the flow path cross-sectional area near the seat portion of the valve body 114 is small. This is caused by an increase in the flow rate of the fuel and a decrease in pressure at the tip of the valve body 114 due to a pressure drop caused by a decrease in static pressure due to the Bernoulli effect. Since this differential pressure is greatly affected by the flow path cross-sectional area of the seat portion, the differential pressure increases when the displacement amount of the valve body 114 is small, and the differential pressure decreases when the displacement amount is large. Therefore, when the valve body 114 starts to open from the closed state and the displacement becomes small and the differential pressure increases, the valve opening of the valve body 114 is impacted by the idle movement of the mover 102 at the timing when the valve opening operation becomes difficult. Therefore, the valve opening operation can be performed even when a higher fuel pressure is applied. Alternatively, the spring 110 can be set to a stronger force for the fuel pressure range that needs to be operable. By setting the spring 110 to a stronger force, the time required for the valve closing operation described later can be shortened, which is effective for controlling the minute injection amount.

  After the valve body 114 starts the valve opening operation, the mover 102 collides with the fixed core 107. When the movable element 102 collides with the fixed core 107, the movable element 102 rebounds. However, the movable element 102 is attracted to the magnetic core by the magnetic attractive force acting on the movable element 102, and then stops. At this time, since a force is applied to the movable element 102 in the direction of the fixed core 107 by the return spring 112, the amount of displacement of the rebound can be reduced, and the time until the rebound converges can be shortened. Since the rebounding action is small, the time during which the gap between the mover 102 and the fixed core 107 is increased is shortened, and stable operation can be performed even with a smaller injection pulse width.

  The movable element 102 and the valve body 114 that have finished the valve opening operation in this manner are stationary in the valve open state. In the valve open state, a gap is formed between the valve body 114 and the valve seat 118, and fuel is injected. The fuel passes through the center hole provided in the fixed core 107, the upper fuel passage hole provided in the mover 102, and the lower fuel passage hole provided in the mover 102, and flows in the downstream direction. .

  When the energization to the solenoid 105 is cut off, the magnetic flux generated in the magnetic circuit disappears and the magnetic attractive force disappears. When the magnetic attractive force acting on the mover 102 disappears, the valve body 114 is pushed back to the closed position in contact with the valve seat 118 by the load of the spring 110 and the force of the fuel pressure.

  Further, when the movable element 102 is divided into the movable element 102a and the movable element 102b, the movable element 102b is provided with a collar provided on the outer diameter thereof in a valve-closed state in which the valve body 114 is in contact with the valve seat 118. The part 211 is in contact with the movable element 102 a, and the movable element 102 b is in contact with the upper end surface of the valve body 114 at the contact surface 210. When the movable element 102a opens from the initial position, the movable element 102b also cooperates to perform the valve opening operation.

  Further, the movable element 102a and the movable element 102b are configured to be slidable on the sliding surface 206, and after the valve body 114 comes into contact with the valve seat 118 when the valve body 114 is closed from the valve open state. The movable element 102a is separated from the valve body 114 and the movable element 102b, moves in the valve closing direction, moves for a certain period of time, and then is returned to the initial position of the valve closed state by the return spring 112.

  Since the movable element 102a is separated from the movable element 102b and the valve element 114 at the moment when the valve element 114 is completely opened, the mass of the movable element 102 can be reduced. The energy can be reduced, and the bounce of the valve body 114 caused by the collision of the valve body 114 with the valve seat 118 can be suppressed.

  In a state where the valve body 114 is stationary at the target lift position, that is, in a valve open state, the movable element 102 and the fixed core 107 collide with one or both of the movable element 102 and the fixed core 107 against the annular end surface facing each other. The protrusion part of the part is provided. Further, in the valve open state, the protrusion has a gap between the movable element 102 or the surface of the fixed core 107 other than the protrusion of the movable element 102 or the fixed core 207 side. One or more fuel passages in which the fluid can move in the outer diameter direction and the inner diameter direction of the protrusions are provided. Due to the effects of the protrusions and the fuel passage, the squeeze that occurs in the direction that hinders the movement of the mover 102 due to the pressure change in the minute gap between the mover 102 and the fixed core 107 can be reduced. There is an effect that the valve closing delay time until the body 114 is closed can be reduced. In general, martensitic or ferritic stainless steel with good magnetic properties has low material hardness and strength. In martensitic stainless steel, magnetic properties decrease when heat treatment is performed to increase the hardness. There is a case. In order to prevent wear of the protrusion due to the collision between the movable element 102 and the fixed core 107, a plating process such as hard chrome plating may be performed on the end surface provided with the protrusion. In the operation in which the valve body 114 is pushed back to the closed position, the mover 102 moves together while being engaged with the regulating portion 114a of the valve body 114.

  In the fuel injection device of this embodiment, the valve body 114 and the movable element 102 are divided into the moment when the movable element 102 collides with the fixed core 107 when the valve is opened and the moment when the valve element 114 collides with the valve seat 118 when the valve is closed. By producing a relative displacement for a very short time, it is possible to suppress the bounce of the movable element 102 against the fixed core 107 and the bounce of the valve body 114 against the valve seat 118.

  By configuring as described above, the spring 110 urges the valve body 114 in the direction opposite to the direction of the driving force by the magnetic attractive force, and the return spring 112 is opposite to the urging force of the spring 110. The mover 102 is biased in the direction.

  Next, the injection pulse output from the drive device 121 for driving the fuel injection device according to the present invention, the drive voltage applied to both terminals of the solenoid 105 of the fuel injection device, the drive current (excitation current), and the valve body of the fuel injection device The relationship between the displacement 114 (valve element behavior) 114 (FIG. 3) and the relationship between the injection pulse and the fuel injection amount (FIG. 4) will be described.

  When an injection pulse is input to the drive circuit 121, the drive circuit 121 applies a high voltage 301 to the solenoid 105 from a high voltage source boosted to a voltage higher than the battery voltage, and starts supplying current to the solenoid 105. . When the current value reaches the peak current value Ipeak preset in the ECU 120, the application of the high voltage 301 is stopped. After that, the voltage value to be applied is set to 0 V or less, and the current value is reduced like the current 202. When the current value becomes smaller than the predetermined current value 304, the drive circuit 121 performs application of the battery voltage VB by switching so that the predetermined current 303 is maintained.

  The fuel injection device is driven by such a supply current profile. Between the time when the high voltage 301 is applied and the peak current value Ipeak is reached, the movable element 102 starts to be displaced at the timing t31. By using the impact, the displacement of the valve body 114 increases sharply, and then the valve body 114 reaches the position of the target lift before shifting to the holding current 303. After reaching the target lift position, the movable element 102 performs a bounce operation due to the collision between the movable element 102 and the fixed core 107, and the valve element 114 is configured to be relatively displaceable with respect to the movable element 102. 114 is separated from the anchor 102, and the displacement of the valve body 114 is displaced beyond the target lift position. Thereafter, the mover 102 stops at a predetermined target lift position and the valve element 114 also stops at the target lift position by the magnetic attraction force generated by the holding current 303 and the force in the valve opening direction of the return spring 112. A stable valve opening state is obtained.

In the case of a fuel injection device having a movable valve in which the valve body 114 and the mover 102 are integrated, the displacement amount of the valve body 114 does not become larger than the target lift position, and the mover 102 and the valve after reaching the target lift. The displacement amount of the body 114 is equivalent. In the case of a fuel injection device in which the movable element 102 and the valve body 114 are integrated, an integral part (hereinafter referred to as a movable valve) serves as a component of the magnetic circuit to generate a magnetic attractive force, and opens and closes the valve seat 118. It has two functions to perform the valve. When the movable element 102 is divided into the movable element 102a and the movable element 102b, the movable element 102b comes into contact with the upper end surface of the valve element 114 and stops after the valve element 114 reaches the valve closing position. The mover 102a moves away from the valve body 114 in the valve closing direction. After the movable element 102a moves for a certain time, the return spring 112 returns the movable element 102a to the initial position of the valve closed state. Since the movable element 102a is separated from the movable element 102b and the valve element 114 at the moment when the valve element 114 is completely opened, the mass of the movable element 102 can be reduced. Energy can be reduced, and bounce of the valve body 114 caused by the valve body 114 colliding with the valve seat 118 can be suppressed. Further, it is preferable that the mass of the movable element 102b is smaller than the mass of the movable element 102a. Due to this effect, the impact force caused by the collision of the valve body 114 with the valve seat 118 can be reduced. Therefore, the bounce of the valve body 114 caused by the collision of the valve body 114 with the valve seat 118 can be suppressed. Unintentional injection after the seat 118 contacts can be suppressed. Next, the relationship between the injection pulse width T and the fuel injection amount will be described with reference to FIG. Under the condition that the injection pulse width Ti does not reach a certain time, the magnetic attractive force acting on the mover 102 does not exceed the force of the set spring 110 acting on the mover 102, so the valve body 114 does not open, Fuel is not injected. Further, even when the magnetic attractive force acting on the mover 102 exceeds the set spring load, the mover 102 cannot move through the gap 201 which is the run-up section, and the injection pulse is stopped. Even if the magnetic attractive force acting on the valve and the inertial force in the valve opening direction of the mover 102 become smaller than the force by the set spring 110, fuel is not injected. Under the condition where the injection pulse width Ti is short, for example, 401, the valve body 114 is separated from the valve seat 118 and starts to lift, but since the valve body 114 starts to close before reaching the target lift position, The injection amount decreases with respect to the alternate long and short dash line 330 extrapolated from the straight line region 420. Further, at the pulse width of the point 402, the valve closing is started immediately after reaching the target lift position, and the locus of the valve body 114 becomes a parabolic motion. Under this condition, the kinetic energy in the valve opening direction of the valve element 114 is large, and the magnetic attraction force acting on the mover 102 is large. The injection amount increases. With the injection pulse width at point 403, valve closing is started at timing t343 when the bounce amount of the mover 102 after reaching the target lift becomes maximum. At this time, the repulsive force when the movable element 102 collides with the fixed core 107 acts on the movable element 102, and the valve closing delay time from when the injection pulse is turned off until the valve body 114 is closed is reduced. The injection amount is smaller than the one-dot chain line 330. Point 404 is a state in which valve closing starts at timing t35 immediately after the bounce of the mover 102 and the bounce of the valve body 114 converge. Under the condition that the injection pulse Ti is larger than the point 404, the injection pulse width Ti increases. Accordingly, since the valve closing delay time increases approximately linearly, the fuel injection amount increases linearly. In the region from the start of fuel injection to the pulse width Ti indicated by point 404, even if the valve body 114 does not reach the target lift or the valve body 114 reaches the target lift, the valve body 114 bounces. Since it is not stable, the injection amount varies.

  In order to reduce the minimum injection amount that can be controlled by the ECU 120, an area in which the fuel injection amount linearly increases in accordance with the increase in the injection pulse width Ti is increased, or the injection pulse width Ti is smaller than 404. It is necessary to correct the injection amount in a non-linear region where the relationship between the pulse width Ti and the injection amount is not linear. In the general driving current waveform as described with reference to FIG. 3, the bounce of the valve body 114 generated by the collision between the movable element 102 and the fixed core 107 is large, and by starting the valve closing in the middle of the bounce of the valve body 114, Non-linearity occurs in the region of the short injection pulse width Ti up to the point 404, and this non-linearity causes the minimum injection amount to deteriorate. Therefore, in order to improve the nonlinearity of the injection amount characteristic under the condition that the valve body 114 reaches the target lift, it is necessary to reduce the bounce of the valve body 114 that occurs after reaching the target lift position. Also. Since there is a variation in the behavior of the valve body 114 due to the dimensional tolerance, the timing of contact between the movable element 102 and the fixed core 107 differs for each fuel injection device, and the collision speed between the movable element 102 and the fixed core 107 varies. The bounce of the valve body 114 varies for each individual fuel injection device, and the individual variation of the injection amount increases.

  Subsequently, FIGS. 5 to 13 will be described. FIG. 5 is a graph showing the relationship between the injection pulse width Ti and the individual variation in the injection amount caused by the component tolerance of the fuel injection device. FIG. 6 is a diagram showing the relationship of the displacement amount of the valve body 114 to the individual variation of the injection amount in FIG. 5 and the relationship between the displacement amount of the valve body 114 and the time at each injection pulse width. FIG. 7 is a diagram showing the relationship among the injection pulse width output from the drive device, the drive current, the displacement amount of the valve body 114, and the mover displacement amount according to the change of time. In the figure of the valve body displacement amount in FIG. 7, the individual valve body start timing and the valve closing completion timing are different, and the valve body displacement amount in the fuel injection device of the conventional structure that does not perform the preliminary operation are described. . FIG. 8 is a diagram showing details of the drive device 121 and the ECU 120 of the fuel injection device. FIG. 9 shows injection pulse widths Ti, drive currents, current differential values, current second-order differential values of three fuel injection devices having different operation timings of the valve body 114 due to the influence of variation in dimensional tolerance in one embodiment of the present invention, It is the figure which showed the relationship between valve body displacement amount, needle | mover displacement amount, and time. FIG. 10 shows an injection pulse, a drive current supplied to the fuel injection device, an operation timing of switching elements 805, 806, and 807 of the drive device, a voltage between terminals of the solenoid 105, a valve body 114 and It is the figure which showed the displacement amount of the needle | mover 102, the needle | mover acceleration, and the relationship of time. FIG. 11 shows the drive current supplied to the solenoid 105, the displacement amounts of the valve bodies of three individuals 1, 2, and 3 whose valve closing behavior differs depending on the dimensional tolerance of the fuel injection device 840, the enlarged view of the voltage VL1, and the voltage VL1. It is the figure which showed the relationship of 2nd-order differential value. 12 shows the displacement (referred to as gap x) between the mover 102 and the fixed core 107, the magnetic flux φ passing through the attraction surface between the mover 102 and the fixed core 107, and the solenoid 105 in this embodiment. It is the figure which showed the corresponding relationship with the voltage between terminals Vinj. FIG. 13 shows the first-order differential values of the terminal voltage Vinj, the drive current, and the current in three fuel injectors having different valve opening start timing and valve opening completion timing under the condition that the valve body in the present embodiment reaches the target lift. It is the figure which showed the relationship between the 2nd-order differential value of an electric current, a valve body displacement amount, and time. FIG. 14 is a diagram showing an initial magnetization curve and a return curve of the magnetization curve (BH curve) of the magnetic material used in the magnetic circuit in the first embodiment. FIG. 15 is a diagram illustrating a flowchart of an injection amount correction method for each cylinder in a region where the injection pulse width Ti is small, which is an intermediate lift region where the valve body does not reach the target lift. FIG. 16 shows the injection amount of each cylinder, valve closing completion timing Tb, valve opening start timing Ta ′ and per unit time injected from the fuel injector 840 when the injection pulse width Ti is changed under a certain fuel pressure condition. It is the graph which showed the relationship of the detection information (Tb-Ta ') * Qst calculated | required from flow volume Qst (henceforth a static flow). FIG. 17 is a diagram showing the relationship between the detection information of the individual fuel injection devices 1, 2, and 3 of each cylinder and the injection pulse width Ti. FIG. 18 shows injection pulse width Ti, drive current, terminal voltage Vinj, second-order differential value of voltage VL1, second-order differential value of current, that is, voltage VL2, and valve under the condition of dividing injection performed during one intake / exhaust stroke. It is the graph which showed the amount of displacement of body 114, and time.

  At first. Using FIGS. The relationship between the injection amount at each injection pulse width Ti and the displacement amount of the valve body 114 and the relationship between the individual variation of the injection amount and the displacement amount of the valve body 114 will be described. Individual variations in the injection amount are the effects of dimensional fluctuations due to component tolerances of the fuel injection device, aging deterioration, and fluctuations in environmental conditions, that is, the fuel pressure supplied to the fuel injection device, the battery voltage source of the drive device, and the voltage of the boost voltage source This is caused by fluctuations in the current value supplied to the solenoid 105 caused by individual variations in values, changes in the resistance value of the solenoid 105 accompanying temperature changes, and the like. The amount of fuel injected from the injection hole 119 of the fuel injection device is equal to the total cross-sectional area of the plurality of injection holes determined by the diameter of the injection hole 119 and the pressure loss from the seat portion of the valve body 114 to the injection hole entrance In this case, the injection amount is determined by the cross-sectional area of the flow path between the valve body 114 and the valve seat 118 through which the fuel in the fuel seat portion determined by the displacement amount of the valve body 114 flows. FIG. 5 shows that an individual Qu with a large injection amount and a small injection amount are smaller than an individual Qc in which the injection amount is the median of the design in a region where the injection pulse width is small when a constant fuel pressure is supplied to the fuel injection device. It is the figure which described individual Ql.

  5 and 6, the relationship between the injection amount at each injection pulse width Ti and the displacement amount of the valve body 114 of the individual Qc, which is the design median value, under the condition of the injection pulse width t51 having a certain injection amount will be described. To do. The amount of displacement of the valve body 114 under the condition of the point 501 where the injection pulse width Ti is small becomes a solid line 501, and before the valve body 114 reaches the target lift, the injection pulse width Ti is turned OFF and the valve body 114 is closed. Starting, the trajectory of the valve body 114 is a parabolic motion. Next, at the point 502 where the injection amount becomes larger than the one-dot chain line 530 extrapolated from the linear region where the relationship between the injection pulse width Ti and the injection amount is substantially linear, the displacement amount of the valve body 114 becomes larger than the solid line 601. The valve body 114 does not reach the target lift position, and starts to close as shown by a one-dot chain line 602, and becomes a parabolic motion locus as in the case of the solid line 601. In addition, since the energy supplied to the solenoid 105 is larger in the alternate long and short dash line 602 than in the solid line 601, the valve closing delay time increases, and as a result, the injection amount also increases. Next, at the point 503 where the injection amount becomes smaller than the one-dot chain line 530, after the movable element 102 collides with the fixed core 107, the valve element 114 starts closing at the timing when the bound of the movable element becomes maximum. The locus is as shown by a two-dot chain line 603, and the valve closing delay time becomes smaller than the condition of the one-dot chain line 602, and as a result, the injection amount at the point 503 becomes smaller than the point 502. Further, 606, 605, and 604 show the displacement amount of the valve body 114 at the points 532, 501, and 531 of each of Qu, Qc, and Ql at the injection pulse width Ti of t51 in the figure. When the injection pulse width 601 at the timing t1 is input to the drive circuit, the timing at which the movable element 102 collides with the valve body 114 after the injection pulse is turned on due to the influence of individual differences in the dimensional tolerance of the fuel injection device 640, The valve opening start timing of the valve body 114 varies as t61, t62, and t63. When the same injection pulse width is given to each cylinder, the displacement amount of the valve body 114 at the timing t64 when the individual valve 604 with the earlier valve opening start timing turns off the injection pulse width becomes the largest. Even after the injection pulse width is turned off, the movable element 102 continues to be displaced by the residual magnetic attraction force accompanying the residual magnetic flux due to the influence of kinetic energy and eddy current, and the kinetic energy and magnetic attraction of the movable element 102 are continued. The valve element 114 starts to close at a timing t67 when the force in the valve opening direction is less than the force in the valve closing direction. As shown by the displacements 604, 604, and 606 of the valve body, the individual with the later valve opening start timing has a larger lift amount of the valve body 114, and after the injection pulse width is turned OFF until the valve body 114 is completely closed. The valve closing delay time increases. Therefore, in the intermediate lift region where the valve body 114 does not reach the target lift, the injection amount is determined by the valve opening start timing of the valve body 114 and the valve closing completion timing of the valve body 114, and therefore, the valve opening start of the fuel injection device of each cylinder If the individual variation in timing and valve closing completion timing can be detected or estimated by the drive unit, the lift amount can be controlled by the intermediate lift, and the individual variation in the injection amount can be reduced to stabilize the injection amount even in the intermediate lift region. Can be controlled.

  Next, the valve operation of each individual fuel injection device having the same valve opening start timing and different valve closing completion timing will be described with reference to FIG. Note that, as described above, the valve body displacement amount in FIG. 7 describes individuals having the same valve opening start timing but different valve closing completion timings.

  From FIG. 7, even when the valve opening start timing t73 is the same due to individual variations of the fuel injection device, as shown in the individual 1, individual 2, and individual 3 of the valve body displacement amount, the valve body is affected by the component tolerance. The differential pressure acting on 114 and the load due to the set spring 110 change for each individual, and the maximum displacement amount of the valve element 114 and the valve closing completion timing vary for each individual. In the individual 3 having a small differential pressure acting on the valve body 114, the force in the valve closing direction is smaller than that of the individual 2 having the median differential pressure, and thus the displacement amount of the valve body 114 is large. As a result, since the magnetic gap between the mover 102 and the fixed core 107 is reduced, even when the same current value is supplied, the magnetic attractive force that is the force in the valve opening direction increases, and the valve closing completion timing is , It is delayed as t76 compared to t75 of the individual 2. On the other hand, in the individual 1 in which the differential pressure is larger than that in the individual 2, the displacement amount of the valve body 114 is reduced, and the magnetic gap between the movable element 102 and the fixed core 107 is increased. The force decreases, and the valve closing completion timing is earlier as t74 than the individual 2 t75. Since the effects of individual pressure differential and magnetic attraction force variations appear in the valve closing completion timing, in addition to the valve opening start timing, the valve closing completion timing should be detected for each fuel injection device of each cylinder by the driving device. Thus, it is possible to detect individual variations in the injection amount.

  Further, in the fuel injection device in which the movable element 102 does not perform a preliminary operation before the valve element 114 starts to open, a magnetic attraction force that is a force in the valve opening direction acting on the movable element and a load by the spring 110. And the valve body 114 starts to open at timing t77 in a state where the difference between the force in the valve closing direction, which is the sum of the differential pressures due to the fuel pressure acting on the valve body 114, is small, and then the valve body as in 701 The displacement amount 114 gradually increases. In the region where the displacement amount of the valve body 114 is small, the flow passage cross-sectional area of the seat portion of the valve body 114 is small, so the flow rate of the fuel flowing through the seat portion is high, and the pressure loss of the fuel due to passing through the seat portion is large . If the pressure loss of the fuel near the seat portion is large, the flow rate of the fuel injected from the injection hole 119 becomes slow, so that the shear resistance between the injected fuel and air is reduced, and the droplets of the injected fuel are atomized. It becomes difficult to promote, and droplets with a large particle size (coarse particle size droplets) are likely to be generated. According to the fuel injection device in the present embodiment, since the movable element 102 collides with the valve body 114 and the valve body 114 starts to open, an area where the displacement amount of the valve body 114 is small can be reduced. The particle diameter of the fuel to be injected can be reduced, and droplets having a large particle diameter are hardly generated. As a result, mixing of the injected fuel and air is easily promoted, and since there are few coarse droplets, the homogeneity of the air-fuel mixture at the ignition timing is improved, and further, the fuel is applied to the piston and cylinder wall surfaces. By suppressing the adhesion, it is possible to improve exhaust performance, particularly to suppress unburned particles (PM) and their number (PN). Moreover, fuel efficiency can be improved by forming an air-fuel mixture with good homogeneity.

  Next, using FIG. 8, FIG. 9, and FIG. 10, the configuration of the drive device of the fuel injection device in this embodiment and the variation in the operation of the valve body 114 that causes individual variation in the injection amount are shown in FIG. A method of detecting by the drive device for each injection device will be described.

  The CPU 801 is built in the ECU 120, for example, a pressure sensor attached to a fuel pipe upstream of the fuel injection device, an A / F sensor for measuring the amount of air flowing into the engine cylinder, and the oxygen concentration of exhaust gas discharged from the engine cylinder. An injection for controlling the amount of fuel injected from the fuel injection device in accordance with the operating conditions of the internal combustion engine by taking in signals indicating the state of the engine such as an oxygen sensor and a crank angle sensor for detecting the engine from the aforementioned various sensors The pulse width and injection timing are calculated.

  Further, the CPU 801 calculates a pulse width (that is, an injection amount) and an injection timing of an appropriate injection pulse width Ti according to the operating conditions of the internal combustion engine, and sends the injection pulse width Ti to the fuel injection device drive IC 802 through the communication line 804. Is output. Thereafter, the drive IC 802 switches between energization and non-energization of the switching elements 805, 806, and 807 to supply a drive current to the fuel injection device 840.

  The switching element 805 is connected between a high voltage source higher than the voltage source VB input to the drive circuit and a high voltage side terminal of the fuel injection device 840. The switching elements 805, 806, and 807 are configured by, for example, FETs, transistors, and the like, and can switch between energization and non-energization of the fuel injection device 840. The boosted voltage VH, which is the voltage value of the high voltage source, is 60 V, for example, and is generated by boosting the battery voltage by the booster circuit 814. For example, the booster circuit 814 includes a DC / DC converter or the like, or a coil 830, a switch element 831, a diode 732, and a capacitor 833. The switch element 831 is, for example, a transistor. A diode 835 is provided between the power supply side terminal 890 of the solenoid 105 and the switching element 805 so that a current flows from the second voltage source in the direction of the solenoid 105 and the ground potential 815. A diode 811 is also provided between the power supply side terminal 890 of the solenoid 105 and the switching element 807 so that current flows from the battery voltage source in the direction of the solenoid 105 and the ground potential 815, and the switch element 808 is energized. During this time, no current flows from the ground potential 815 toward the solenoid 105, the battery voltage source, and the second voltage source.

  When the booster circuit 814 includes the coil 830, the switch element 831, the diode 832, and the capacitor 833, when the transistor 831 is energized, the battery voltage VB flows to the ground potential 834 side, but when the transistor 831 is de-energized, The high voltage generated in the coil 830 is rectified through the diode 832 and electric charge is accumulated in the capacitor 833. The switch element 831 is repeatedly energized / de-energized until the boosted voltage VH is reached, and the voltage of the capacitor 833 is increased. The energization / non-energization of the switch element 831 may be configured to be controlled by the IC 802 or the CPU 801.

  The switching element 807 is connected between the low voltage source VB and the high voltage terminal of the fuel injection device. The low voltage source VB is, for example, a battery voltage, and the voltage value is about 12 to 14V. The switching element 806 is connected between the low voltage side terminal of the fuel injection device 840 and the ground potential 815. The drive IC 802 detects the current value flowing through the fuel injection device 840 by the current detection resistors 808, 812, and 813, and switches between energization / non-energization of the switching elements 805, 806, and 807 according to the detected current value. The desired drive current is generated. The current detection resistors 808, 812, and 813 are high-precision resistors that have a small resistance value and a small individual variation in resistance value from the viewpoint of improving current detection accuracy, reliability, and suppressing heat generation. A shunt resistor should be used. In particular, since the resistance values of the resistors 808, 812, and 813 are sufficiently smaller than the resistance value of the solenoid 105 of the fuel injection device 840, the influence of the loss generated in the resistors 808, 812, and 813 on the current of the solenoid 105 is not affected. small. Diodes 809 and 810 are provided to apply a reverse voltage to the solenoid 105 of the fuel injection device and to rapidly reduce the current supplied to the solenoid 105. The CPU 801 communicates with the drive IC 802 through the communication line 803. The drive current generated by the drive IC 802 can be switched according to the pressure of the fuel supplied to the fuel injection device 840 and the operating conditions. Further, both ends of the resistors 808, 812, and 813 are connected to an A / D conversion port of the IC 802, and the voltage applied to both ends of the resistors 808, 812, and 813 can be detected by the IC 802. Capacitors 851 and 850 are provided on the Hi side (voltage side) and the ground potential (GND) side of the fuel injection device 840 to protect the input voltage and output voltage signals from surge voltage and noise, respectively. A resistor 852 and a resistor 853 may be provided in parallel with the capacitor 850 downstream of the injection device 840.

  An active low-pass filter 861 including an operational amplifier 821, resistors R83 and R84, and a capacitor C82 is provided between a terminal 880 between the switching element 806 and the resistor 808 and the CPU 801 or the IC 802. An operational amplifier 820, resistors R81 and R82, a capacitor C81, and a terminal 881 between the resistor 852 and the resistor 853 provided on the ground potential (GND) side of the fuel injection device 840 and the CPU 801 or IC 802 are provided. An active low-pass filter 860 is provided. The CPU 801 or the IC 802 is provided with a terminal 871 connected to the ground potential 815 so that the potential difference VL1 between the terminal 881 and the ground potential 815 can be detected by the CPU 801 or the IC 802 through the active low-pass filter 860. , A terminal y80 is provided. Further, the resistor 852 and the resistor 853 are set to be larger than the resistance value of the solenoid 105 of the fuel injection device 840, so that a current is efficiently supplied to the solenoid 105 when a voltage is applied to the fuel injection device 840. . Further, by setting the resistance value of the resistor 852 larger than that of the resistor 853, the voltage VL between the ground potential (GND) side terminal of the fuel injection device 840 and the ground potential can be divided. As a result, the detected voltage can be set to VL1, and the withstand voltage of the A / D conversion port of the operational amplifier 821 and the CPU 801 can be reduced, so that a circuit necessary for inputting a high voltage is not required. The time of the voltage generated in the terminal voltage Vinj and the voltage VL can be detected.

  Further, a terminal y81 may be provided so that the potential difference VL2 between the terminal 880 on the fuel injection device 840 side of the resistor 808 and the ground potential 815 can be detected by the CPU 801 or the IC 802 through the active low-pass filter 861. The CPU 801 is provided with a terminal y82 connected to the battery voltage VB so that the battery voltage VB can be monitored by the CPU 801.

  Next, a method for detecting the valve opening start timing of the valve body 114 in the first embodiment of the present invention will be described with reference to FIG. FIG. 9 shows three individuals with different operation timings, but important ones as the operation timing include a valve opening start timing and a valve closing completion timing. Further, a change that occurs in the current flowing through the solenoid 105 can be detected by the driving device by detecting the voltage VL2.

  From FIG. 9, the boosted voltage VH is applied to the solenoid 105 until the current supplied to the solenoid 105 reaches the peak current Ipeak. Thereafter, the voltage boosting voltage VH in the negative direction is applied, or a voltage of 0 V is applied, so that a voltage cutoff period T2 in which the current value decreases as in 901 and the current decreases for a certain time is provided. In the movable element 102, after the boost voltage VH is applied to the solenoid 105, a magnetic attraction force that is a force in the valve opening direction that acts on the movable element 102 is a spring 110 that is a force in the valve closing direction that acts on the movable element 102. When the force by is exceeded, the mover 102 is displaced in the valve opening direction, and the idle running operation is performed. Thereafter, the valve body 114 starts to be displaced at timings t91, t92, and t93 at which the individual movable elements 102 of the fuel injection device 840 come into contact with the valve body 114, and fuel is injected from the injection holes 119. Before the valve body 114 starts to open, the peak current Ipeak or the boosted voltage application time tp and the voltage cutoff period T2 may be adjusted so that the timing t91 when a constant voltage is supplied from the battery voltage source.

  In the fuel injection device 840 of the present embodiment, the movable element 102 collides with the valve body 114 after the idling operation, so that the force due to the fuel pressure that has been applied only to the valve body 114 so far is transmitted via the valve body 114. Since it acts on the movable element 102, the acceleration of the movable element 102 changes greatly at the valve opening start timing of the valve element 114. Between the mover 102 and the fixed core 107 is a main path through which the magnetic flux of the magnetic circuit composed of the fixed core 107, the mover 102, the nozzle holder 101, the housing 103, and the solenoid 105 passes. As the acceleration changes, the magnetic flux passing between the mover 102 and the fixed core 107 changes, a change occurs in the induced electromotive force, and a change occurs in the slope of the current value. In order to detect the current gradient, that is, the timing at which the current differential value changes, the ECU detects the timing at which the second-order differential value of the current becomes the maximum value. Thereby, the valve opening start timing can be detected for each fuel injection device 840 of each cylinder. Further, in the section from the timing t91 after the constant voltage is supplied from the battery voltage source to the valve opening start timing of the valve body 114, the switching elements 805, 806, and 807 are not switched between energization and non-energization, By eliminating the electrical change of the drive current and reducing the time change of the current, the effect of facilitating detection of the change in acceleration caused by the collision of the movable element 102 with the valve body 114 can be obtained, and the valve opening is started. Timing detection accuracy can be improved.

  Here, a terminal y81 for measuring the voltage VL2 may be provided in the CPU 801 in order to detect a time change of the current flowing through the solenoid 105 by the driving device. Since the resistance value of the resistor 808 is known, the current flowing through the solenoid 105 by detecting the voltage VL2 from the relationship of Ohm's law V = R · I (the voltage V is the product of the resistance R and the current I). Can be detected. Further, even when the resistance value of the resistor 808 varies due to individual variation or resistance temperature change, according to the method of detecting the timing at which the second-order differential value of the current becomes maximum, the second-order voltage VL2 is detected. Even when the maximum value of the differential value fluctuates, the time for the voltage VL2 to become the second-order differential value does not change, so that the valve opening start timing can be detected with higher accuracy, and the detection robustness is high. The voltage VL2 is connected to the A / D conversion port of the CPU 801 via the active low-pass filter 861. The CPU 801 detects the time at which the second-order differential value of the current becomes the maximum value by detecting the digital signal obtained by A / D converting the voltage VL2 by the CPU 801, thereby detecting the valve opening start timing of the valve element 114. can do. Further, the drive device may store the valve opening start delay time from when the injection pulse is turned on until the valve opening start timing is reached. In addition, in the valve opening start timing, when the current that has been decreasing until now starts increasing, the valve opening start timing can be detected as the timing at which the current differential value exceeds a certain threshold value. However, due to the configuration of the fuel injection device 840 and the drive device, even if the current does not turn from increasing to decreasing at the valve opening start timing, the second-order differential value of the current becomes the maximum value after the injection pulse is turned on. By detecting the valve opening start delay time until it becomes, it becomes possible to detect the valve opening start timing with high accuracy.

  Note that the voltage cutoff period T2 is not always essential, but by applying a negative boosted voltage VH or 0V, it is easy to detect a change in the current flowing through the solenoid 105 for reasons described later.

  In addition, when all the voltages VL2 during the period when the ejection pulse is ON are detected by the driving device, the inflection point generated in the current due to the energization / non-energization of the current switching elements 805, 806, 807 is the second floor of the voltage VL2. There is a case of erroneous detection as a differential value. In this case, by setting the acquisition period of the voltage VL2 to a period 903 in which the switching operation of the switching elements 805, 806, and 807 is not performed, the valve opening start timing at which the movable element 102 collides with the valve body 114 is set. Can be detected with high accuracy. The time t98a at which data acquisition in the period 903 is started is set later than the time t91 that is the end timing of the voltage cutoff period T2, and the time 98b at which data acquisition in the period 903 is stopped is the time t98 at which the injection pulse is turned off. It is better to set it earlier. As a trigger for starting time t98a, the timing of the start of the injection pulse and the energization / non-energization timing of the switching elements 805 and 806 may be used. When the energization / non-energization timing of the switching elements 805 and 806 is used as a trigger at time t98a, the energization / non-energization information of the switching elements 805 and 806 may be transmitted to the CPU 801 through the communication line 803.

  When the start of the injection pulse is used as a trigger, since the injection pulse is generated inside the CPU 801, the time t98a can be accurately controlled. On the other hand, when the timing at which the switching elements 805 and 806 are de-energized is used as a trigger for starting time t98a, the peak current value Ipeak is changed due to the resistance change accompanying the temperature change of the solenoid 105 and the fluctuation of the boost voltage VH. Even when the boosted voltage application time Tp to reach the fluctuating value can be obtained with certainty, the detection accuracy of the valve opening start timing can be improved.

  As described above, in order to detect the valve opening start timing of the valve body 114, it is desirable to detect the second-order differential value of the voltage VL2 for detecting the current flowing through the solenoid 105 by the driving device. In the case of performing second-order differentiation processing with a high degree of differentiation processing, if noise or the like is superimposed on the voltage VL2 before processing, the differentiation value may diverge when the differentiation processing is performed. There is a possibility of misdetecting the timing of the maximum value after the step differentiation process. In order to cope with this problem, an active low-pass filter 861 including an operational amplifier 821, a resistor R83, a resistor R84, and a capacitor C82 may be configured between the terminal 880 of the fuel injection device 840 and the terminal y81 of the CPU 801. In this embodiment, the valve opening start timing is detected by utilizing the fact that the acceleration of the movable element 102a is changed by the collision of the movable element 102a with the valve body 114 and the current and the voltage VL2 of the solenoid 105 are changed. ing. The changes that occur in the current and voltage VL2 of the solenoid 105 at this time have a lower frequency than the noise superimposed on the signal of the current and voltage VL2. Therefore, by passing the active low-pass filter 861 between the terminal 880 for regulating the voltage VL2 and the CPU 801, high-frequency noise generated in the current and the voltage VL2 can be reduced, and the valve opening start timing is detected. Accuracy can be increased.

  The cut-off frequency fc1 of the active low-pass filter 861 can be expressed as the following equation (1) using the values of the resistor R82 and the capacitor C81. The switching timing of the switching element 831 for configuring the switching elements 805, 806, 807 and the second voltage source and the value of the second voltage source differ depending on the configurations of the fuel injection device and the driving device. The frequency of the noise generated in is different. Therefore, the design values of the resistor R82 and the capacitor C81 are preferably changed for each specification of the fuel injection device 840 and the drive circuit. Further, when the low-pass filter is configured with an analog circuit, the CPU 801 does not need to perform filtering processing for removing digital high-frequency noise, so that the calculation load on the CPU 801 can be reduced. Alternatively, the signal of the voltage VL1 may be directly input to the CPU 801 or the IC 802 and digitally filtered. In this case, it is not necessary to use the operational amplifier 820, the resistor R81, the resistor R82, and the capacitor C81, which are components of the analog low-pass filter, so that the cost of the driving device can be reduced. The low-pass filter described above may be a primary low-pass filter including a resistor connected to the terminal 880 and a capacitor arranged in parallel with the resistor. When the primary low-pass filter is used, the cost of the driving device can be reduced because the two components of the resistor and the operational amplifier can be reduced compared to the configuration using the active low-pass filter. Moreover, the calculation method of the cut-off frequency of the primary low-pass filter can be calculated by Expression (1) in the case of using the active low-pass filter. Further, as a configuration of the low-pass filter, it is possible to configure a low-pass filter having a second or higher time using a coil and a capacitor. In this case, since a low-pass filter can be configured without a resistor, there is a merit that power consumption is small compared to the case where an active low-pass filter and a primary low-pass filter are used.

fc1 = 1 / (2 · π · R84 · C82) (1)
Note that the detection of the current of the solenoid 105 for detecting the valve opening start timing may be performed by measuring the voltage across the resistor 813. However, when measuring the voltage across the resistor 813, the number of terminals for measuring the voltage and the required A / D conversion ports also increase compared to the voltage VL2 for measuring the potential difference from the ground potential 815. This leads to an increase in the cost of the driving device, and the processing load on the CPU 801 or IC 802 for A / D conversion of the voltage signal increases. In addition, at the voltage VL2, the switching element 831 is repeatedly energized / de-energized at high speed for charge accumulation in the capacitor 833 for restoring the voltage value of the boosted voltage VH, which is the output of the booster circuit 814. In this case, a high-frequency noise component may be superimposed on the voltage across the resistor 813 that is the path on the power source side of the fuel injection device 840. By setting the current measurement point to the voltage VL2 positioned on the ground potential side of the solenoid 105 of the fuel injection device 840, high-frequency noise generated upstream of the fuel injection device 840 is attenuated by the coil of the solenoid 105, so that the voltage VL2 The valve opening start timing can be detected with high accuracy using the maximum of the second-order differential value.

  Next, referring to FIG. 2, FIG. 8, and FIG. 10, the configuration of the drive circuit in the first embodiment and the switching element for generating the drive current flowing in the fuel injection device under the condition for detecting the valve opening start timing are shown. The switching timing will be described.

  First, at timing t101, when the injection pulse width Ti is input from the CPU 801 to the drive IC 802 through the communication line 804, the switching elements 805 and 806 are energized, and the boosted voltage VH is applied to both ends of the solenoid 105 to drive. Current is supplied to the solenoid 105 and the current increases rapidly. Thereafter, a magnetic flux is formed inside the magnetic circuit with the disappearance of the eddy current generated in the magnetic circuit, and the magnetic flux passes between the fixed core 107 and the movable element 102 to act on the movable element 102. Magnetic attraction increases. The mover 102 starts to lift at timing t102 when the sum of the magnetic attractive force acting on the mover 102 that is the force in the valve opening direction and the force of the return spring 112 exceeds the load of the spring 110 that is the force in the valve closing direction. . At this time, when the mover 102 moves in the valve opening direction, shear resistance (viscous resistance) is generated between the mover 102 and the nozzle holder 101, and the shear resistance force is movable in the valve closing direction opposite to the movement direction. It acts on the child 102. However, by securing a passage cross-sectional area between the mover 102 and the nozzle holder 101, the shear resistance acting on the mover 102 can be reduced. Further, since the shear resistance force acting on the movable element 102 is sufficiently smaller than the magnetic attractive force that is the force in the valve opening direction acting on the movable element 102, the movable element 102 starts to lift after the movable element 102 starts to lift. The acceleration increases. When the switching elements 805 and 806 that have been energized are de-energized at the timing t103 when the drive current reaches the peak current value Ipeak that is given to the ECU in advance, the solenoid 105 and the ground potential are increased from the boosted voltage VH through which the current has been flowing. Since no current flows in the path toward 815, the voltage of the ground potential (GND) side terminal of the fuel injection device 840 increases due to the back electromotive force due to the inductance of the fuel injection device 840, and the ground potential 815 of the driving device, the diode 809, a fuel injection device 840, a diode 810, a resistor 812, and a current path of the boosted voltage VH are formed, and the current returns to the boosted voltage VH side of the booster circuit 814. The voltage across the solenoid 105 has a negative direction. The boosted voltage VH is applied and the drive supplied to the solenoid 105 Stream is rapidly reduced as 1002.

  By setting the timing t103 when the switching elements 805 and 806 are de-energized as the timing when the drive current exceeds the peak current value Ipeak, a change in resistance value due to a temperature change of the solenoid 105 and a change in voltage value of the boost voltage VH can be achieved. Even if it occurs, the energy required to open the valve body 114 is stably secured, and the valve opening start timing caused by the time fluctuation until the peak current Ipeak is reached due to the change in environmental conditions. Can be used as a component of translation, and changes in the current waveform and the timing of the valve operation can be suppressed.

  Further, the timing t103 at which the switching elements 805 and 806 are de-energized may be set by the boosted voltage application time Tp after the injection pulse Ti is turned on. Since the setting resolution of the peak current Ipeak is determined by the resistance value and accuracy of the resistors 808 and 813 used for current detection, the minimum value of Ipeak resolution that can be set by the driving device is limited by the resistance of the driving device. On the other hand, when the timing t103 at which the switching elements 805 and 806 are deenergized is controlled by the boost voltage application time Tp, the setting resolution of the boost voltage application time Tp is not limited by the resistance of the driving device, and the Since it can be set according to the clock frequency, the time resolution can be reduced as compared with the case of setting with the peak current Ipeak, and the boost voltage application time Tp or the peak current value Ipeak is stopped with higher accuracy. Since the timing can be corrected, the correction accuracy of the injection amount of the fuel injection device of each cylinder can be improved.

  Further, the time of the voltage cutoff period T2 in which the switching elements 805 and 806 are de-energized may be stored in advance in the drive device and changed according to operating conditions such as fuel pressure. When the voltage cutoff period T2 ends, the switching elements 806 and 807 are energized, and the battery voltage VB is applied to the solenoid 105. At this time, the current value of the target value Ih1 of the drive current is set to a value higher than the current at the end of the voltage cutoff period T2, such as 1004, so that the switching element 806 until the target current is reached. Continue to be ON. At this time, after the timing t105 when the switching elements 806 and 807 are energized, the charge accumulated in the capacitors 851 and 852 is discharged, so that the drive current increases as 1003. Thereafter, current is supplied to the solenoid 105 by application of the battery voltage, the displacement amount of the mover 102 increases, and the current starts to decrease at timing t105 due to the induced electromotive force generated as the magnetic gap decreases, and timing t106. , The movable element 102 collides with the valve body 114. At this time. When the movable element 102 collides with the valve body 114, the differential pressure due to the fuel pressure acting on the valve body 114 acts on the movable element 1-2 via the valve body 114, so that the acceleration of the movable element 102 changes greatly. To do. As the induced electromotive force changes as the acceleration of the mover 102 changes, the slope of the drive current changes. Since the movable element 102 collides with the valve body 114 and the switching elements 806 and 807 are energized at the valve opening start timing of the valve body 114, the change in the inter-terminal voltage value Vinj is small, and the boost voltage VH Since a lower battery voltage VB is applied, the change in current accompanying the application of the voltage becomes gentle. Therefore, a slight change in the induced electromotive force due to the collision of the mover 102 with the valve body 114 is reduced. The change can be detected by the driving device. In addition, since the current is rapidly reduced from the peak current value Ipeak and the current value at the valve opening start timing of the valve body 114 is reduced, the magnetic field generated in the magnetic circuit is reduced, and the magnetic flux density is caused accordingly. As a result, the magnetic flux density on the end surface of the movable element 102 on the fixed core 107 side is hard to be saturated, and as a result, the movable element 102 collides with the valve element 114 and the valve element 114 starts to open the valve. The change is more easily detected as a current time change, that is, a change in the current gradient. By setting the values of the peak current Ipeak and the voltage cutoff period T2 so that the valve element 114 starts to open during the period when the switching elements 806 and 807 are energized and the battery voltage VH is applied to the solenoid 105. The valve opening start timing of the valve body 114 can be detected with high accuracy.

  Further, the displacement amount of the valve body 114 shown in FIG. 10 describes a displacement profile of the valve body 114 when the fuel pressure supplied to the fuel injection device 840 is small, medium, and large. In the fuel injection device 840 according to the first embodiment, the movable element 102 does not receive the force due to the fuel pressure acting on the valve body 114 until the valve body 114 starts to open, so even if the fuel pressure is different. The profile of the movable element 102 until the movable element 102 collides with the valve body 114 does not change, and the valve opening start timing t106 of the valve body 114 does not change. Therefore, the detection of the valve opening start timing t106 of the valve body 114 is detected under certain operating conditions such as idle operation when the engine is started, and stored in the drive device, thereby changing the operating conditions such as fuel pressure. Even in this case, the detection information of each cylinder stored in the drive device can be used. Therefore, the driving device A for converting the analog voltage signal of the potential difference VL2 between the voltage across the resistor 813 for detecting the drive current for detecting the valve opening start timing or the ground potential 815 of the resistor 808 into a digital signal. Since the frequency of using the / D conversion port can be reduced, the processing load on the CPU 801 and the IC 802 can be reduced. As described above, if the valve opening start timing is detected under a certain operating condition for each fuel injection device 840 of each cylinder, the detection accuracy can be ensured even when the operating condition such as the fuel pressure changes.

  In addition, in order to monitor the voltage value of the battery voltage VB of the battery voltage source, the CPU 801 is provided with a terminal y82 which is an A / D conversion port for A / D converting the voltage and detecting it as a digital signal by the driving device. ing. The voltage of the battery voltage VB drops due to the operation of the in-vehicle device connected to the battery voltage source, and the fluctuation thereof is large. In-vehicle devices are, for example, cell motors used when starting an engine, air conditioners such as air conditioners, lights (headlights, brake lamps), and electric power steering. Further, the alternator is started in accordance with the voltage drop to charge the battery voltage source. Therefore, the voltage VL2 or the voltage across the resistor 813 when the battery voltage VB monitored by the CPU 801 falls below a certain fluctuation range of the voltage value set in the driving device is detected, and the valve opening start timing is detected. It is good to comprise so that it may detect. With the above configuration, in the condition for detecting the valve opening start timing, the battery voltage VB changes depending on the operation of the in-vehicle device, and the current is affected when the change timing of the battery voltage is close to the valve opening start timing. Therefore, it is possible to suppress the possibility that the time when the current second-order differential value for detecting the valve opening start timing becomes the maximum value is shifted, and the valve opening start timing can be detected stably.

  In addition, the median value of the voltage value under the condition for detecting the valve opening start timing also changes depending on the deterioration of the battery voltage source. Thereby, even when the median value of the battery voltage VB when the battery voltage source is not used has changed over time, the valve opening start timing can be detected with high accuracy.

  Further, since the ferrite magnetic material having a high saturation magnetic flux density used for the magnetic circuit member of the fuel injection device 840 in this embodiment has a lower material hardness than the austenitic metal, the valve element 114 of the mover 102 is used. There is a case where the plating process is performed on the collision surface with the fixed core 107 and the collision surface with the fixed core 107. Since the mover 102 opens the valve at high speed without receiving a force due to the fuel pressure and collides with the valve body 114, the total number of revolutions of the engine increases and the number of times the fuel injection device 840 is driven increases. The collision surface 210 of the movable element 102 with the valve body 114 may be worn out. In particular, when it is desired to improve the homogeneity of the mixture of fuel and air in order to suppress the total amount of unburned particles PM including soot and the number PN, fuel injection during one intake / exhaust stroke is performed several times. Although the method of dividing into two parts is effective, the number of injections increases even if the mileage is the same as when the divided injections are performed, so that wear deterioration of the collision surface 210 occurs. easy. When the wear deterioration occurs, the gap 201 between the contact surface 205 of the valve element 114 with the movable element 102a and the collision surface 210 of the movable element 102a in the closed state increases, and the movable element 102 becomes the valve element 114. The moving distance necessary for the collision increases, and the valve opening start timing of the valve body 114 is delayed. The fuel injection device 840 for each cylinder is re-detected at predetermined intervals and stored in the drive device according to the number of times of driving the fuel injection device 840, the time, or the value of the travel distance measuring device attached to the vehicle. Even when the number of times of driving of the fuel injection device 840 is increased by performing split injection by updating the information on the valve opening start timing of the valve, it responds to a change in the valve opening start timing due to deterioration wear on the collision surface Therefore, it is possible to control the injection amount with high accuracy.

  Further, under the condition that the switching elements 805 and 806 are energized and the boosted voltage VH in the positive direction is applied to the solenoid 105, the use of the boosted voltage VH reduces the charge accumulated in the capacitor 833 so far. The voltage value of the voltage VH decreases. At this time, in order to restore the voltage of the boosted voltage VH to the initial voltage value preset in the CPU 801 or IC 802, if the voltage value of the boosted voltage VH falls below the set threshold voltage, the charge is stored in the capacitor 833. Therefore, there is a case where the switching element 831 of the booster circuit 814 is repeatedly energized / deenergized at a high frequency to restore the voltage value of the boosted voltage VH. The change in the induced electromotive force caused by the acceleration change of the mover 102 due to the collision of the valve body 114 with the valve body 114 and the valve body 114 starting to open has little influence on the voltage VL2 and the voltage across the resistor 812. The change in acceleration of the movable element 102 when the valve element 114 starts to open is measured under the condition that the boosted voltage VH is applied. It is difficult to detect at the second voltage across. Further, when performing an operation for restoring the voltage value of the boosted voltage VH, it is necessary to repeat energization / non-energization of the switching element 831 of the booster circuit 814 at a high cycle, which causes high-frequency noise due to switching. In some cases, noise is superimposed on the voltage VL2 for detecting the valve opening start timing of the valve body 114 or the voltage across the resistor 812, which adversely affects the detection accuracy of the valve opening start timing.

  From FIG. 9, after supplying the injection pulse width Ti, the switching elements 805 and 806 are energized, the boosted voltage VH is applied to the solenoid 105, and the negative boosted voltage VH is constantly applied after reaching the peak current value Ipeak. After the time has elapsed and the current value has fallen sharply as 901, a constant voltage that is the battery voltage VB is applied from the battery voltage source, and the valve body 114 is moved to the target lift at the timing of supplying the constant voltage from the battery voltage VB. It is preferable to adopt a configuration of an applied voltage that reaches

  Next, a method for detecting the valve closing delay time, which is the time from when the injection pulse is turned OFF to when the valve body 114 closes, will be described.

  Further, when the valve body 114 and the mover 102 are closed from the open state, the time of the voltage generated in the voltage VL that is the potential difference between the ground potential (GND) side terminal of the fuel injection device 840 and the ground potential 815. In order to detect the change by the CPU 801 or the IC 802, resistors 852 and 853 are provided between the ground potential side (GND) side terminal of the fuel injection device 840 and the ground potential 815. By setting the resistance values of the resistors 852 and 853 to be larger than the resistance value of the solenoid 105, current can efficiently flow through the solenoid 105 when the battery voltage VB and the boost voltage VH are applied. Further, by setting the resistance value of the resistor 852 to be larger than the resistance value of the resistor 853, the voltage of VL1 that is a potential difference with respect to the ground potential 815 of the resistor 853 can be reduced, and the operational amplifier 821 and the CPU 801 Since the voltage value of the withstand voltage required for the A / D conversion port of the terminal can be reduced, the voltage generated in the inter-terminal voltage Vinj and the voltage VL without requiring a circuit or an element necessary for inputting a high voltage. Can be detected. The voltage VL1 obtained by dividing the voltage VL is input to the CPU 801 or the A / D conversion port mounted on the IC 802 via the active low-pass filter 860. The high-frequency noise component generated in the voltage VL1 can be reduced by passing the signal of the voltage VL1 through the active low-pass filter 860, and the valve body 114 starts to close from the open state and contacts the valve seat 117. The change in the acceleration of the movable element 102 generated in the above can be detected by the voltage VL1 as a change in the induced electromotive force, and can be detected as a digital signal by the IC 802 or the CPU 801. As a result, differentiation processing can be easily performed. At this time. A potential difference between the terminal y80 that passes through the active low-pass filter 860 and is input to the A / D conversion port of the CPU 801 and the ground potential 815 is referred to as a voltage VL3.

  Next, using FIG. 2, FIG. 8, FIG. 11, and FIG. 12, the description of the operation of the drive circuit in the first embodiment and the individual variation in the injection amount of the fuel injection device 840 along with the individual variation in the valve opening start timing. The detection principle of the valve closing completion timing for calculating the valve closing delay time, which is the time from when the injection pulse is turned OFF until the valve body 114 contacts the valve seat 118, which causes variation, will be described. Since the time change of the terminal voltage Vinj also occurs in the voltage VL and the voltage VL1, the voltage change in FIG. 11 is equivalent to the time change of the voltage VL1 detected by the CPU 801. Further, the movable element 102b is in contact with the movable element 102a at an end surface 204 provided on the movable element 102a, and the movable element 102a and the movable element 102b can be relatively displaced.

  Further, as shown in FIG. 11, when the injection pulse width Ti is turned OFF, the disappearance of the magnetic flux starts from the vicinity of the solenoid 105 due to the influence of the eddy current generated in the magnetic material of the magnetic circuit, and is generated in the mover 102a and the mover 102b. The valve body 114 starts to close at a timing when the magnetic attraction force is reduced and the magnetic attraction force falls below the force in the valve closing direction acting on the valve body 114 and the movable element 102a and the movable element 102b. The magnitude of the magnetic resistance of the magnetic circuit is inversely proportional to the cross-sectional area in each path through which the magnetic flux passes and the magnetic permeability of the material, and is proportional to the magnetic path length through which the magnetic flux passes. Compared to a magnetic metal having a high saturation magnetic flux density, the magnetic permeability of the gap between the mover 102 and the fixed core 107 is a vacuum magnetic permeability μ0 = 4π × 10 −7 [H / m], and the magnetic material Since the magnetic permeability of this metal is very small, the magnetic resistance increases. The magnetic permeability μ of a magnetic material is determined by the characteristics of the BH curve (magnetization curve) of the magnetic material due to the relationship of B = μ · H, and varies depending on the magnitude of the internal magnetic field of the magnetic circuit. The magnetic permeability becomes low, the magnetic permeability increases with increasing magnetic field [], and the magnetic permeability decreases when a certain magnetic field is exceeded. When the valve body 114 is displaced from the valve opening position, a gap x is generated between the mover 102 and the fixed core 107. Therefore, the magnetic resistance of the magnetic circuit increases, the magnetic flux that can be generated in the magnetic circuit decreases, and the mover 102 The magnetic flux passing through the suction surface on the fixed core 107 side end surface is also reduced. When the magnetic flux generated in the magnetic circuit of the solenoid 105 changes, an induced electromotive force is generated according to Lenz's law. In general, the magnitude of the induced electromotive force in the magnetic circuit is proportional to the rate of change of the magnetic flux flowing through the magnetic circuit (the first-order differential value of the magnetic flux). Assuming that the number of turns of the solenoid 105 is N and the magnetic flux generated in the magnetic circuit is φ, the terminal voltage Vinj of the fuel injection device is expressed by an induced electromotive force term −N · (dφ as shown in Expression (2). / Dt) and the sum of the resistance component R of the solenoid 105 generated by Ohm's law and the product of the current i flowing through the solenoid 105.

Vinj = −N · (dφ / dt) + R · i (2)
When the valve body 114 comes into contact with the valve seat 118, the movable element 102a is separated from the movable element 102b and the valve body 114, but by the spring 110 that has been acting on the movable element 102a through the valve body 114 and the movable element 102b until now. The force in the valve closing direction, which is the force due to the load and the fuel pressure acting on the valve body 114, no longer acts, and the mover 102 a is biased in the valve opening direction by the force of the return spring 112. That is, the direction of the force acting on the movable element 102a at the moment when the valve body 114 completes the valve closing changes from the valve closing direction to the valve opening direction, so that the acceleration of the movable element 102a changes.

  The relationship between the gap x generated between the mover 102 and the fixed core 107 and the magnetic flux φ passing through the attraction surface can be regarded as a first-order approximation relationship in a very short time. When the gap x increases. The distance between the mover 102 and the fixed core 107 is increased, and the magnetic resistance is increased. The magnetic flux that can pass through the end face on the fixed core 107 side of the mover 102 is reduced, and the magnetic attractive force is also reduced. The attractive force acting on the mover 102 can theoretically be derived from the equation (3). From equation (3), the attractive force acting on the movable element 102 is proportional to the square of the magnetic flux density B of the attractive surface of the movable element 102 and proportional to the attractive area S of the movable element 102.

Fmag = (B ² · S) / (2 · μ ₀ ) (3)
From Equation (2) and FIG. 12, there is a correspondence between the voltage Vinj between the terminals of the solenoid 105 and the first-order differential value of the magnetic flux φ passing through the attraction surface of the mover 102. In addition, since the gap x, which is the distance between the end surface of the mover 102 on the fixed core 107 side and the end surface of the fixed core 107 on the mover 102 side, changes, the area of the space between the mover 102 and the fixed core 107 increases. Since the magnetic resistance of the magnetic circuit changes and, as a result, the magnetic flux that can pass through the attraction surface of the mover 102 changes, it can be considered that the gap x and the magnetic flux φ have a first-order approximation relationship in a very short time. it can. When the gap x is small, the area of the space between the mover 102 and the fixed core 107 is small, so the magnetic resistance of the magnetic circuit is small, and the magnetic flux that can pass through the attraction surface of the mover 102 increases. On the other hand, when the gap x is large, the area of the space region between the mover 102 and the fixed core 107 is large, so the magnetic resistance of the magnetic circuit is large, and the magnetic flux that can pass through the attracting surface of the mover 102 decreases. To do. From FIG. 12, the first-order differential value of the magnetic flux has a corresponding relationship with the first-order differential value of the gap x. Further, the first-order differential values of the terminal voltage Vinj and the voltage VL2 correspond to the second-order differential value of the magnetic flux φ, and the second-order differential value of the magnetic flux φ is the second-order differential value of the gap x, that is, the acceleration of the movable element 102. Equivalent to. Therefore, in order to detect a change in the acceleration of the mover 102, it is necessary to detect the second-order differential value of the voltage Vinj or the voltage VL between the terminals. For this purpose, the voltage VL is divided and the voltage VL2 is converted into the A / V of the CPU 801. It may be input to the D conversion port.

  From FIG. 11, when the injection pulse width Ti is stopped, that is, the energization to the solenoid 105 is stopped and the valve body 114 starts to be displaced from the maximum displacement position, the profile of the voltage VL2 changes. Further, the voltage VL2 changes in accordance with the amount of displacement of the mover 102 that moves in conjunction with the valve body 114. The larger the gap x between the mover 102 and the fixed core 107, the larger the magnetic resistance, so the residual magnetic flux decreases, and as a result, the voltage VL2 gradually approaches 0V.

  Further, the movable element 102a is separated from the movable element 102b and the valve element 114 at the moment when the valve element 114 comes into contact with the valve seat 118, so that the movable element 102a has been acting on the movable element 102a through the movable element 102b and the valve element 114 so far. The force in the valve closing direction no longer acts, and the movable element 102a receives the force in the valve opening direction of the return spring 112, and the direction of the force acting on the movable element 102a changes from the valve closing direction to the valve opening direction. Therefore, the change in the acceleration of the mover 102a can be detected by the minimum value of the second-order differential value of the voltage VL2. Note that the voltage detection point may be the terminal voltage Vinj, the voltage VL2, or the voltage VHL obtained by subtracting the voltage VL from the voltage VH between the terminal 890 and the ground potential. When the valve closing completion timing of the valve body 114 is detected with the voltage VL, an inflection point may occur at the timing when the discharge of the capacitor 851 ends in the voltage VL after stopping the injection pulse Ti. When the time difference between the timing at which this inflection point occurs and the valve closing completion timing is small, the inflection point generated by the capacitor 851 may interfere with the voltage VL at the valve closing completion timing, resulting in a detection error. By detecting the voltage VHL or the terminal voltage Vinj, the inflection point generated by the capacitor 851 can be canceled, so that the detection accuracy of the valve closing completion timing can be improved.

  After the injection pulse width Ti is stopped, the movable element 102a and the movable element 102b are displaced from the target lift position in conjunction with the valve body 114, and the voltage VL at this time gradually decreases from the value of the positive boost voltage VH to 0V. Asymptotically. When the movable element 102a is separated from the valve element 114 and the movable element 102b after the valve element 114 is closed, the force in the valve closing direction that has been acting on the movable element 102a through the valve element 114 and the movable element 102b so far, that is, the spring The load due to 110 and the force due to fuel pressure disappear, and the load of the return spring 112 acts on the movable element 102a as the force in the valve opening direction. When the valve body 114 reaches the valve closing position and the direction of the force acting on the movable element 102a changes from the valve closing direction to the valve opening direction, the second-order differential value of the voltage VL, which has been gently decreased until now, is increased. Roll over. By detecting the minimum value of the second-order differential value of the voltage VL with the drive circuit, it is possible to detect individual variations in the displacement amount of the valve body 114 with high accuracy. Further, the value of the voltage VL due to the displacement of the movable element 102a and the movable element 102b from the valve opening position is the resistance value determined by the wire diameter and the number of windings of the solenoid 105, the specification of the magnetic circuit, the material of the magnetic material ( It varies depending on the inductance determined by the electrical low efficiency and BH curve), the design value of the target lift of the valve body 114, the current value at the timing when the injection pulse width Ti is stopped, and the tolerance variation of the dimensions and setting values explained above Is greatly affected by. In the method for detecting the valve closing delay time based on the second-order differential value of the voltage VL, since the change point of the acceleration of the movable element 102a and the movable element 102b is detected as a physical quantity, variation in design value, tolerance, and environmental condition (current value) ), The valve closing completion timing can be detected with high accuracy, and the valve closing delay time, which is the time from when the injection pulse is turned OFF until the valve body 114 is closed, can be detected.

  In order to detect the time from when the injection pulse width T is stopped until the valve body 114 is closed, the second-order differentiation is performed on the voltage VL1 obtained by dividing the voltage Vinj input to the IC 802 or the CPU 801 or the voltage VL. And the exact valve closing completion timing can be detected by detecting the timing when the second-order differential value becomes the minimum as the timing when the valve element 114 is closed. Further, in the pre-processing for detecting the inter-terminal voltage Vinj or the voltage VLI obtained by dividing the voltage VL, the operational amplifier 820, the resistor R81, the resistor R82, and the capacitor C81 are provided between the terminal 881 of the fuel injection device 840 and the terminal y80 of the CPU 801. An active low-pass filter 860 composed of Changes in the voltage Vinj between the terminals, the voltage VL, and the voltage VL1 caused by the change in the acceleration of the movable element 102a when the valve body 114 is closed are lower in frequency than the noise superimposed on the voltage signal. Therefore, by passing an active low-pass filter between the terminal 881 for measuring the voltage VL1 and the CPU 801, high-frequency noise generated in the terminal voltage Vinj, the voltage VL, and the voltage VL1 can be reduced. The detection accuracy of the valve completion timing can be increased.

  Further, the cut-off frequency fc2 of the active low-pass filter 860 can be expressed by the following equation (4) using the values of the resistor R84 and the capacitor C82. The switching timing of the switching element 831 for configuring the switching elements 805, 806, 807 and the second voltage source and the value of the second voltage source differ depending on the configurations of the fuel injection device and the driving device. The frequency of the noise generated in is different. Therefore, it is preferable to change the design values of the resistor R84 and the capacitor C82 for each specification of the fuel injection device 840 and the drive circuit. Further, when the low-pass filter is configured by an analog circuit, it is not necessary to perform the filtering process digitally by the CPU 801, so that the calculation load on the CPU 801 can be reduced. Alternatively, the signal of the voltage VL1 may be directly input to the CPU 801 or the IC 802 and digitally filtered. In this case, it is not necessary to use the operational amplifier 820, the resistor R81, the resistor R82, and the capacitor C81, which are components of the analog low-pass filter, so that the cost of the driving device can be reduced. The low-pass filter described above may be a primary low-pass filter including a capacitor arranged in parallel with a resistor arranged in series with the terminal 853. When the primary low-pass filter is used, the cost of the driving device can be reduced because the two components of the resistor and the operational amplifier can be reduced compared to the configuration using the active low-pass filter. Further, the calculation method of the cutoff frequency of the primary port-pass filter can be calculated by the equation (4) when the active low-pass filter 860 is used. This cutoff frequency fc2 may be configured to be different from the value of the active low-pass filter fc1 for detecting the valve opening start timing.

  Further, as a configuration of the low-pass filter, it is possible to configure a low-pass filter having a second or higher time using a coil and a capacitor. In this case, since a low-pass filter can be configured without a resistor, there is a merit that power consumption is small compared to the case where an active low-pass filter and a primary low-pass filter are used.

fc2 = 1 / (2 · π · R82 · C81) (4)
The voltage measurement point for detecting the valve closing completion timing can be the inter-terminal voltage Vinj, but the inter-terminal voltage Vinj is a high frequency generated by the switching element 831 of the booster circuit of the fuel injection device 840. Noise is generated. In the inter-terminal voltage Vinj, the voltage profile after the ejection pulse Ti is stopped is reversed in positive and negative as compared with the voltage VL, and gradually approaches the voltage 0V from the boosted voltage VH in the negative direction. Therefore, in order to detect the valve closing completion timing, it is necessary to detect the maximum value of the second-order differential value of the inter-terminal voltage Vinj, but in order to detect with high accuracy, a low-pass is required to reduce switching noise. Since it is necessary to set a large time constant for the filter, an error may occur between the timing of contact between the valve body 114 and the valve seat 118 and the valve closing completion timing detected by the second-order differential value of the detected terminal voltage Vinj. is there. Since this error becomes a variation in detection and may be a constraint for performing the minute injection amount control, the position at which the valve closing completion timing is measured is not the inter-terminal voltage Vinj but the ground potential side terminal of the fuel injection device 840 and the ground potential (GND). It is desirable to measure the voltage VL, which is the potential difference from

  The signal input to the CPU 801 or the IC 802 is triggered by the injection pulse width Ti, and the signal of the voltage VL2 is set for a predetermined time after the injection pulse width Ti is stopped. Take it in. With such a configuration, the data point sequence of the voltage VL2 input to the CPU 801 or the IC 802 can be suppressed to the minimum necessary for detecting the valve closing completion timing. Therefore, the storage capacity of the CPU 801 and the memory of the IC 802 And the calculation load can be reduced. In addition, if the voltage differentiation process is performed at the timing of switching from the boosted voltage VH to the battery voltage VB, the timing of repeating energization / non-energization of the switching elements 805, 806, and 807, that is, the timing at which the voltage change becomes steep, Therefore, there is a possibility that the valve closing completion timing may be erroneously detected when the valve closing completion timing at which the valve element 114 contacts the valve seat 118 is detected by the second-order differential value of the voltage VL2. However, it is possible to prevent erroneous detection of the valve opening completion timing by determining the voltage detection period by the CPU 801 or the IC 802.

  The voltage detection resistor 816 may be a shunt resistor with high resistance accuracy. In the driving device of the fuel injection device 840, the voltage at both ends of the voltage detection resistors 812, 813, 808, and 816 provided in the driving circuit is diagnosed by the IC 802 or the CPU 801 in order to measure current or voltage. If the resistance value differs for each individual with respect to the resistance value set in the CPU 801, an error occurs in the voltage value estimated by the IC 802, and the drive current supplied to the solenoid 105 of the fuel injection device 840 is changed to each cylinder. The fuel injection device 840 fluctuates and the injection amount variation increases.

  In addition, when the voltage Vinj between the terminals of the fuel injection device 840 is small at the valve closing position where the valve body 114 and the valve seat 118 are in contact with each other, the change in the voltage value due to the change in the acceleration of the mover 102 becomes relatively small. A method of increasing the load of the spring 110 and shortening the valve closing delay time is effective so that the valve closing position is reached under the condition that the voltage Vinj between the terminals of the solenoid 105 is high. Further, as the fuel pressure supplied to the fuel injection device 840 increases, the force due to the fuel pressure acting on the valve body 114 and the mover 102 increases, and therefore the valve closing delay time decreases. For example, the individual variation of each cylinder at the valve closing completion timing at which the valve element 114 contacts the valve seat 118 is detected under the condition that the fuel pressure is high and the fuel pressure supplied to the fuel injection device 840 in each cylinder is the same. And good. Due to this effect, the residual magnetic flux generated in the magnetic circuit at the valve closing completion timing becomes larger than in the condition where the fuel pressure is low, and the speed when the valve body 114 collides with the valve seat 118 increases. The moment when the valve element 114 contacts the valve seat 118, the change in the acceleration of the mover 102 due to the movement of the mover 102 away from the valve element 114 increases, and the change in the induced electromotive force also increases. Alternatively, it becomes easy to detect the valve closing completion timing with the voltage Vl2 differential value. Further, under the condition where the fuel pressure supplied to the fuel injection device 840 is high and the engine load is large, the injection amount injected during one intake stroke increases, and the pressure pulsation of the pipe attached upstream of the fuel injection device 840 increases. The fuel pressure supplied to the fuel injection device 840 may fluctuate due to the influence. In this case, the valve closing completion timing may be detected under conditions such as idle operation where the engine load is small and the injection amount of each cylinder is the same.

  In addition to the CPU 801 and the IC 802, a microcomputer for detecting the voltage VL2 and processing data may be provided. When the CPU 801 detects the voltage VL1 and the voltage VL2 and performs data processing, it is necessary to perform A / D conversion on the data at a high sampling rate and perform differential processing. When the signal from another sensor is captured When interrupt processing occurs or when the calculation load on the CPU 801 is high, it may be difficult to detect and differentiate the voltages VL1 and VL2. Therefore, the voltage VL1 and the voltage VL2 are detected by a microcomputer provided in addition to the CPU 801, masking processing and differentiation processing are performed, the second-order differential values of the voltages VL1 and VL2 are calculated, and the second-order differential value of the voltage is minimum. By detecting the maximum timing as the valve closing completion timing and the valve opening start timing, and having the function of storing them in the microcomputer, the calculation load of the CPU 801 and the IC 802 can be reduced, and the valve opening completion timing can be reliably detected. Since it can be performed, the correction accuracy of the injection amount can be improved. This microcomputer is provided with a communication line that can communicate with the CPU 801 or the IC 802, and the CPU 801 stores the fuel pressure information taken from the pressure sensor by the CPU 801 and the detection information of the valve closing completion timing transmitted from the microcomputer. It is good to comprise so that it may. With such a configuration, it is possible to more reliably detect the valve opening start timing and the valve closing completion timing, so that the injection amount of each cylinder can be controlled more accurately.

  As a first alternative means for detecting the valve closing completion timing, a method of detecting the inflection point of the league current flowing in the coil 105 after the injection pulse Ti is stopped is conceivable. When the injection pulse Ti is stopped while the drive current is supplied to the coil 105, the switching elements 805, 806, and 807 are de-energized, the negative boost voltage VH is applied to the coil 105, and the drive current is rapidly reduced. Is done. When the drive current reaches around 0A, the voltage that has been generated by the counter electromotive voltage disappears so far, and the current does not flow to the path that has been fed back to the boost voltage VH side. The voltage application is stopped, but the coil current slightly flows in the coil 105. At this time, since the switching elements 805, 806, and 807 are all OFF, the league current flows from the coil 105 to the ground potential 815 side through the resistors 852 and 853. Therefore, in order to detect this leakage current, the voltage across the resistor 852 or 853 is measured, or a shunt resistor is provided in the path between the coil 105 and the ground potential 815, and the voltage across the voltage is measured. Can be considered. Alternatively, when the current reaches around 0 A and the application of the negative boost voltage VH is stopped, the switching element 806 is turned on, and the leakage current is caused to flow from the resistor 808 to the ground potential 815 side. By measuring the voltage across the resistor 808, which is a highly accurate shunt resistor, and performing differential processing on the voltage, the inflection point of the leakage current can be detected, and the valve closing completion timing of the valve body 114 can be detected. .

  Further, as a second alternative means for detecting the valve closing completion timing, which is the moment when the valve body 114 comes into contact with the valve seat 118, an acceleration pickup is provided on the injector of each cylinder or on the engine side where the injector is fixed. A method of detecting the valve closing completion timing by detecting an impact when the valve body 114 collides with the valve seat 118 or a vibration caused by a water hammer caused by suddenly stopping the fuel injection is also considered. It is done. In this case, in order to accurately detect the valve closing completion timing of each cylinder, the accelerometer is attached to a cylindrical portion on the side surface of the injector housing and fixed to the accelerometer, and the accelerometer is attached to the accelerometer with a mounting screw or the like. By using and fixing to the housing, it is possible to easily detect the vibration accompanying the valve closing completion timing of the injector. Further, in the method using the acceleration pickup, the movable element 102 can simultaneously detect the valve opening completion timing when it collides with the fixed core 107. On the other hand, the acceleration pickup and the output voltage are set to each injector. An amplifier for amplification and two wirings of a voltage signal and a GND line are required. In addition, in order to perform detection with high accuracy, it is necessary to increase the sampling rate for accurately processing the high-frequency vibration waveform obtained by the accelerometer so that a high-performance A / D converter can be used. Necessary.

  Further, as a third alternative means for detecting the valve closing completion timing that is the moment when the valve body 114 contacts the valve seat 118, a pressure sensor provided in the rail piping upstream of the injector for detecting knocking or an engine A method of using a sensor for detecting knocking attached to the device can be considered. In a state where fuel is being injected from the injector, the pressure in the rail pipe decreases, and the pump performs a pressurizing operation by the amount of the pressure decrease so that the target fuel pressure is reached by the pump attached upstream. When the valve body 114 collides with the valve seat 118 from the valve open state and reaches the valve closing completion timing, the pressure reduction of the fuel pipe upstream of the injector stops, and the valve closing is completed by detecting the inflection point of the pressure. A method for detecting timing is conceivable. Further, since a sensor used for detecting knocking is generally a vibration pickup that detects vibration, vibration at the time of valve closing caused by the collision of the valve body 114 with the valve seat 118 accompanying the valve closing completion timing of the injector It is possible to detect the vibration at the time of valve opening that occurs when the mover 102 collides with the fixed core 107, and it is possible to detect the timing of completion of opening and closing. When this is used, the engine is running at low speed and the load is small so that the opening and closing timings of other cylinders and the opening and closing timings detected by the vibrations during combustion do not match It is good to detect the valve opening completion timing and the valve closing completion timing under conditions.

  In a normal engine, a command value from an A / F sensor (air-fuel ratio sensor) is detected by the CPU 801, and the injection pulse width is finely adjusted for each fuel injection device of each cylinder even under the same operating conditions. Under the condition for detecting the valve closing completion timing, fine adjustment of the injection pulse width based on the command value from the A / F sensor is stopped, and the valve opening start and valve closing completion timing are supplied under the condition that the same injection pulse width is supplied. Should be detected. By doing so, it is possible to reduce the influence of fluctuations other than individual variations accompanying the valve operation of the fuel injection device 840, such as variations in inflow air when detecting the valve closing start timing and the valve closing completion timing, It is possible to accurately detect variations in the fuel injection device 840 for each fuel injection device of each cylinder in the valve opening start timing and the valve closing completion timing.

  Further, when the injection pulse width Ti is stopped and the valve body 114 is closed from the opened state, the valve body 114 comes into contact with the valve seat 118 after the valve body 114 or the movable element 102 starts to close. Thus, until the valve closing is completed, the switching operation of the driving device may be controlled so that the switching elements 805, 806, and 807 of the driving device are not switched between energization and non-energization. With the above configuration, high-frequency measurement noise caused by switching the switching elements 805, 806, and 807 is not superimposed on the inter-terminal voltage Vinj or VL voltage. Can be improved.

  Next, a method for detecting the valve opening completion timing, which is the timing at which the valve body 114 reaches the target lift, will be described with reference to FIG. The driving current, the first-order differential value of the current, the second-order differential value of the current, and the displacement amount of the valve body 114 shown in FIG. Three individual profiles of the fuel injection device 840 having different operation timings are described. As shown in FIG. 13, by first applying the boosted voltage VH to the solenoid 105, the current is rapidly increased to increase the magnetic attractive force acting on the mover. Thereafter, the valve opening start timings of the valve bodies 114 of the individual 1, individual 2, and individual 3 that are fuel injection devices of the respective cylinders are reached by timing t1303 when the driving current reaches the peak current value Ipeak and the voltage cutoff period T2 ends. It is preferable to set the peak current value Ipeak or the peak current arrival time Tp and the voltage cut-off period T2 so that. Under the condition in which the battery voltage VB is continuously applied and a constant voltage value 1301 is supplied, the change in the applied voltage to the solenoid 105 is small, so that the mover 102 starts to lift from the valve closing position, A change in magnetoresistance associated with a reduction in the gap with the fixed core 107 can be detected as a change in induced electromotive force. When the valve body 114 and the mover 102 start to lift, the gap between the mover 102 and the fixed core 107 is reduced, so that the induced electromotive force is increased and the current supplied to the solenoid 105 is gradually reduced as 1303. To decrease. Since the change in the induced electromotive force accompanying the change in the gap becomes small at the timing when the mover 102 reaches the fixed core 107, that is, the timing when the valve element 114 reaches the target lift (hereinafter referred to as valve opening completion timing), The value increases gradually as 1304. Although the magnitude of the induced electromotive force is affected by the current value in addition to the gap, since the change in current is small under the condition that a voltage lower than the boosted voltage VH is applied like the battery voltage VB, the gap It is easy to detect the change in the induced electromotive force due to the change in the current.

  In order to detect the timing when the valve body 114 reaches the target lift for the individual 1, individual 2, and individual 3 of each cylinder of the fuel injection device 840 described above, First-order differentiation is performed, and timings t113, t114, and t115 at which the first-order differential value of the current becomes 0 may be detected as the valve opening completion timing.

  In addition, in the configuration of the drive unit and the magnetic circuit in which the induced electromotive force generated due to the change in the gap is small, the current may not necessarily decrease due to the change in the gap, but by reaching the valve opening completion timing, Since the slope, that is, the differential value of the current changes, the valve opening completion timing can be detected by detecting the maximum value of the second-order differential value of the current detected by the driving device, and the restrictions on the magnetic circuit, inductance, resistance value, and current Therefore, the valve opening completion timing can be detected stably, and the correction accuracy of the injection amount can be improved.

  Further, the detection of the valve opening completion timing is performed by detecting the valve opening completion timing described in the separate structure of the valve body 114 and the movable element 102 even in the configuration of the movable valve in which the valve body 114 and the movable element 102 are integrated. It can be detected by the same principle.

  Here, the BH characteristic of the magnetic material used in the magnetic circuit of the fuel injection device 840 in the first embodiment is shown in FIG. From FIG. 14, the BH curve of the magnetic material has a non-linear relationship between the input magnetic field and the magnetic flux density. When an increasing magnetic field is applied to the non-magnetized magnetic material, the magnetic material begins to be magnetized and the magnetic flux density Increases until the saturation magnetic flux density Bs is reached. In this process, there is a region H1 where the gradient between the magnetic field and the magnetic flux density is large, and a region H2 where the gradient between the magnetic field and the magnetic flux density is small. Further, when the magnetic field is decreased after reaching the saturation magnetic flux density Bs, the phenomenon that the magnetic material is magnetized is delayed in time, thereby drawing a curve different from the initial magnetization curve. Since the fuel injection device 840 often repeatedly applies a positive magnetic field, a hysteresis minor loop is often drawn between the initial magnetization curve and the return curve. Therefore, under the condition for detecting the valve opening start timing and the valve opening completion timing, the current is increased until the peak current Ipeak is reached, and the magnetic attraction force necessary for displacing the valve body 114 is generated in the mover 102. It is preferable to reduce the magnetic attractive force acting on the mover 102 by providing a period T2 during which the drive current is rapidly reduced before the valve opening start timing and the valve opening completion timing. The drive current supplied to the solenoid 105 of the fuel injection device 840 is supplied to the solenoid 105 under a condition where the drive current is higher than the current value necessary for holding the valve body 114 in the valve open state, such as the peak current value Ipeak. The current value increases, and as shown in FIG. 14, the current value is often located in a region H2 where the gradient between the magnetic field and the magnetic flux density is small, and the magnetic flux density is near saturation. In the first embodiment, after generating the magnetic attractive force necessary for opening the movable element 102, the boosted voltage VH in the negative direction is applied during the period T2, and the current is rapidly decreased. The drive current at the valve opening start timing and the valve opening completion timing is reduced, and the gradient of the magnetic field and the magnetic flux density can be increased as compared with the gradient of the magnetic field and the magnetic flux density under the condition of the peak current value Ipeak. The change in the acceleration of the movable element 102 at the timing when the valve 114 starts to open can be more easily detected as the maximum value of the second-order differential value of the voltage VL2. Similarly, at the valve opening completion timing, the valve body 114 starts to be displaced, and the change in the magnetic resistance accompanying the reduction in the gap between the mover 102 and the fixed core 107 can be more easily detected as the change in the induced electromotive force. effective.

  As described above, when detecting the valve opening start or completion timing, it is not always necessary to apply the boosted voltage VH or 0 V in the negative direction after increasing the current until the peak current Ipeak is reached. By doing so, it becomes possible to detect the valve opening start or completion timing with better accuracy.

  Further, in detecting the valve opening completion timing, there is a time after the elapse of a certain time given to the driving device from the time when the peak current value Ipeak is reached or the time when the application of the negative boost voltage VH is completed. It is preferable to detect only the current value in the period and perform the first-order differentiation process of the current value. By adopting such a configuration, the current value changes rapidly at the timing when the boosted voltage VH is turned ON / OFF. Therefore, the threshold value previously given to the drive device at a time that is not the valve opening completion timing is Can be suppressed, and the detection accuracy of the valve opening completion timing can be improved. Note that the peak current is set so as not to reach the target current value Ih1 set in advance in the IC 802 during the period in which the voltage value 1301 is supplied from the battery voltage source VB after the application of the negative boost voltage VH is stopped. The period Thb during which the value Ipeak and the negative boost voltage VH are applied may be adjusted. Due to this effect, if the drive current reaches the target current value Ih1 before the valve body 114 reaches the target lift, the drive device is controlled to keep the current Ih1 constant. Can pass through the zero point repeatedly, so that it is possible to solve the problem that the change in the induced electromotive force cannot be detected by the drive current.

  Further, from the state where the constant voltage value 1102 is applied, the application of the negative boost voltage VH or the voltage is stopped (application of 0 V), and the current value reaches the current 704 in FIG. By repeating ON / OFF of the voltage VB, the switching elements 605, 606, and 607 are controlled so that the current 703 is obtained. The time from when the injection pulse width Ti is turned on until the current value Ih1 is reached varies depending on the individual difference of the valve body 114 and the variation in the valve opening completion timing accompanying the change in the fuel pressure. The magnetic attraction force when the injection pulse width Ti is stopped depends greatly on the value of the drive current when the injection pulse width Ti is turned off. When the drive current is large, the magnetic attraction force increases and the valve closing delay time increases. To increase. On the contrary, if the drive current when the injection pulse width Ti is turned OFF is small, the hourly suction force becomes small and the valve closing delay time is reduced. As described above, the current value at the timing at which the injection pulse width Ti is turned OFF is desirably the same current 703 for each individual under the condition for detecting the completion of valve opening. The timing of applying the boosted voltage VH in the direction or stopping the application of the voltage may be controlled by the time after the injection pulse width Ti is turned ON or the time after the peak current value Ipeak is reached.

  In the method for detecting and estimating the variation in the injection amount of each cylinder in the first embodiment, the time from the supply of the injection pulse width Ti to the completion of the valve opening is defined as the valve opening delay time for each fuel injection device 840 of each cylinder. The deviation value from the median value of the valve opening delay time stored in advance and given to the CPU 801 is calculated, and the correction value of the injection pulse width Ti after the next injection is calculated according to the deviation value, and the valve opening delay time is calculated. The injection pulse width Ti may be corrected for each fuel injection device 840 of each cylinder based on the detected information. By correcting the injection pulse width Ti based on the detection information of the valve opening delay time, it is possible to reduce individual variations in the injection amount caused by variations in the valve opening delay time due to variations in tolerance.

  Next, a control method in the case of performing an intermediate lift operation using information on the valve opening completion timing of the fuel injection device 840 detected in the present embodiment will be described. Under the condition in which the valve body 114 does not reach the target lift and the intermediate lift operation is performed, the individual variation in the injection amount is determined by the variation in the valve opening start timing and the valve closing completion timing. However, when the fuel injection device is not driven in a state where the drive device and the fuel injection device are connected, the intermediate lift operation for detecting the valve opening start timing and the valve closing completion timing is not performed. When an intermediate pulse is output by outputting the injection pulse width to obtain the calculated injection amount, the injection amount variation with respect to the assumed injection amount increases depending on the fuel injection device of each cylinder, and mixing Qi fuel may be rich or lean and in some cases misfire. Therefore, before the first intermediate lift operation is performed, it is necessary to detect the valve opening completion timing and estimate the valve opening start timing under the condition that the valve body 114 reaches the target lift. In this case, the valve opening completion timing is estimated by multiplying the valve opening delay time for each fuel injection device of each cylinder stored in the drive device by the correction coefficient using the detection waveform of the valve opening completion timing. good. Further, in order to accurately estimate the valve opening start timing, the correlation coefficient between the valve opening completion timing and the valve opening start timing needs to be high, so that the fuel acting on the valve body 114 that affects the valve opening completion timing. The valve opening start timing may be estimated from information on the valve opening delay time under the low twist pressure condition in which the differential pressure due to the pressure is small.

  Next, a method for correcting the injection amount at the intermediate lift will be described with reference to FIGS. 4, 15, 16, and 17. FIG. 15 is a flowchart of the injection amount correction in the region of the injection pulse width smaller than the point 402 in FIG.

  When performing the intermediate lift operation for the first time, the drive device does not obtain the detection information of the valve opening start and valve opening completion timing during the intermediate lift operation of each cylinder, so that the valve element 114 reaches the target lift. The valve opening completion time and the valve opening start timing are estimated by multiplying the valve opening delay time and the valve closing delay time detected for each fuel injection device 840 of each cylinder by a correction coefficient given to the CPU 801 in advance. The actual injection period (Tb−Ta ′) at the intermediate lift calculated from the valve start timing Ta ′ and the valve closing completion timing Tb is calculated, and the preset value and the actual injection period (Tb−Ta ′) given to the CPU 801 in advance. It is preferable to perform the intermediate lift operation by correcting the injection pulse width Ti by the deviation value of. Further, from FIG. 15, the flow rate Qst (per unit time) injected from the fuel injection device 840 under the condition that the actual injection period (Tb−Ta ′) as the detection information and the valve body 114 are stationary at the target lift position. Hereinafter, the relationship between the value obtained by multiplying by (static flow), (Tb−Ta ′) · Qst, and the injection amount is functionalized and set in advance in the CPU 801 of the driving device. From FIG. 16, for example, the relationship between the injection amount and (Tb−Ta ′) · Qst can be obtained as a first-order approximation relationship. From FIG. 17, detection information (Tb−Ta ′) · Qst at each injection pulse width is acquired, and the coefficient of each cylinder is determined from the detection information based on the relationship between the injection pulse width Ti and the detection information (Tb−Ta ′) · Qst. To decide. The relationship between the detection information (Tb−Ta ′) · Qst and the injection pulse width Ti can be expressed by, for example, a first-order approximation, and the coefficients of the functions of the individuals 1, 2, and 3, a1, b1, a2, The coefficients b2, a3, and b3 can be calculated from the detection information. The CPU 801 can detect detection information at two points with different ejection pulse widths Ti, and calculate a coefficient. According to the flowchart described above, when the required injection amount is determined by the CPU 801, the injection amount at the intermediate lift can be corrected by correcting the injection pulse width Ti for each cylinder. Control becomes possible.

  Next, a control method of the fuel injection device 840 for obtaining detection information at the intermediate lift will be described with reference to FIG.

  In the fuel injection system constituted by the fuel injection device and the drive device in this embodiment, the fuel pressure supplied to the fuel injection device and the injection pulse Ti differ between the valve opening start timing and the valve closing completion timing in the intermediate lift condition. It is necessary to acquire multiple times depending on the condition. However, when the detection information at the intermediate lift is not obtained, the injection amount at the intermediate lift is estimated from the valve opening completion timing and the valve closing completion timing under the condition that the valve body 114 reaches the target lift, and the intermediate lift operation is performed. There is a need to do. In this case, the deviation value from the target injection amount becomes large, the ratio of air to fuel to be sucked (air-fuel ratio) becomes rich and lean, and a lot of unburned substances are discharged by unstable combustion, Exhaust performance may deteriorate and in some cases misfire may occur.

  From FIG. 18, the injection during one intake / exhaust stroke is divided into a plurality of times, and a certain amount of injection is performed under the condition that the valve body 114 whose variation in the injection amount of each cylinder is known reaches the target lift. Before performing the injection with the intermediate lift, it is possible to detect the valve opening start timing and the valve closing completion timing during the intermediate lift operation. At this time, the integral value of the displacement amount of the valve body 114 corresponds to the injection amount, and is set so that the injection amount at the intermediate lift is smaller than the injection amount when the valve body 114 reaches the target lift. Good. As a result, since most of the injection amount during one intake / exhaust stroke is determined by the injection amount under the condition of reaching the target lift, misfire can be suppressed even if the injection amount at the intermediate lift deviates from the target value. effective.

  The injection for obtaining the detection information of the valve closing completion timing under the condition of the intermediate lift may be performed once or a plurality of times in one intake / exhaust stroke. The valve closing completion timing for correcting the injection amount by performing the intermediate lift operation a plurality of times during one intake / exhaust stroke and using different injection pulse widths Ti in the first intermediate lift operation and the second intermediate lift operation. A plurality of detection information can be obtained simultaneously. Further, when the detection information of the valve opening start timing has already been obtained, it is not necessary to use the waveform of the second injection shown in FIG. What is necessary is just to use the current waveform suitable for performing. According to the above method, it is possible to obtain the detection information of the valve closing completion timing in the intermediate lift while maintaining the combustion stability, so that the fuel injection device of each cylinder in the intermediate lift condition in a short time Individual variations can be corrected and minute fuel injection can be performed.

  Further, according to the method in the present embodiment, not only the individual variation at the intermediate lift, but also when the valve body 114 is driven under the condition of reaching the target lift, Variation in the injection amount of the injector can be reduced. This is because the individual dispersion of the valve opening completion timing after the valve body 114 starts closing after the injection pulse Ti is stopped is caused by a tolerance fluctuation of a dimension that determines the set spring load and the magnetic attractive force. Therefore, for an individual whose valve closing completion timing is early, the movable element 102 is separated from the fixed core 107, and the valve closing start timing at which the valve body 114 starts closing is also earlier. Therefore, the value obtained by adding the flow rate per unit time at the full lift to the fluctuation time of the valve closing completion timing corresponds to the amount of fluctuation in the injection amount due to individual variation in the valve closing completion timing, and therefore the valve closing completion timing is detected. Thus, the injection amount variation until the valve body 114 reaches the valve closing completion timing from the valve open state can be derived by the ECU. Further, the injection amount that is injected until the valve body 114 reaches the target lift is derived from the inclination of the valve body 114 that can be estimated from the information of the valve opening start timing and the valve opening completion timing of the injectors detected by the ECU. Therefore, it is possible to detect the injection amount variation of the injector of each cylinder together with the injection amount variation estimated from the valve closing completion timing, and to correct the valve by correcting the injection pulse width Ti and the current setting value. It is possible to correct the injection amount under the condition that the body 114 reaches the target lift.

  Further, as shown in FIG. 18, after acquiring information on the valve opening start timing and the valve closing completion timing in the intermediate lift operation, it is preferable to perform the divided injection performed in one intake stroke by the intermediate lift operation. When operating with an intermediate lift, the time required for the valve body 114, the movable element 102a, and the movable element 102b to accelerate in the valve closing direction after the injection pulse Ti is stopped, as compared to when the valve element 114 reaches the target lift and operates. Is short. Accordingly, since the speed of the valve body 114, the movable element 102a, and the movable element 102b at the timing when the valve body 114 contacts the valve seat 118 can be reduced, the movable element 102a is closed in the valve closing direction after the valve body 114 is closed. Thus, the time required for the parabola to return to the position where the return spring 112 comes into contact with the valve body 114 again can be shortened. When the injection pulse of the next injection in the divided injection is applied while the mover 102b is moving, the injection pulse is turned on by the kinetic energy of the mover 102b in addition to the magnetic attractive force acting on the mover 102b. The time until the movable element 102b collides with the valve body 114 is shortened, and the valve opening start timing of the valve body 114 is earlier, and the injection amount varies between the first injection and the second injection. Become. In the present embodiment, the valve opening start delay time and the valve closing completion delay time are stored in the drive device for each fuel injection device of each cylinder, so that divided injection during one intake / exhaust stroke is performed by an intermediate lift operation. As a result, since the injection interval for performing the next injection after the valve body 114 is closed can be reduced, the number of divided injections can be increased, and more precise injection amount control and injection can be performed. Since the timing can be controlled, the homogeneity of the air-fuel mixture can be improved. Further, in the intermediate lift, since the injection amount is small compared to the case where the valve body 114 is driven by reaching the target lift, the penetration force of the spray of the injected fuel can be weakened. The cylinder wall surface adhesion can be suppressed, the number of unburned particles PM including soot and the number PN of unburned particles can be reduced, and the exhaust gas can be made cleaner.

  The configuration of the fuel injection device and the drive device according to the second embodiment of the present invention will be described with reference to FIGS. FIG. 19 is an enlarged view showing a cross section of the drive unit in the valve closing state in which the valve body is in contact with the valve seat in the fuel injection device according to the present embodiment. FIG. 20 is an enlarged view of the longitudinal section of the valve body tip of the fuel injection device. FIG. 21 is an enlarged view showing a section of the drive section in the valve open state for the fuel injection device in the present embodiment. FIG. 22 is an enlarged view of a cross section of the drive unit at the moment when the valve body starts to close from the open state and contacts the valve seat 118. FIG. 23 is a diagram showing the configuration of the driving device in the present embodiment. FIG. 24 shows the relationship between the terminal 2306 and the ground potential 815 for detecting the displacement amount and voltage VL of the second valve element 1907 and the second movable element 1902 when the valve is closed from the maximum lift in the intermediate lift state. It is the figure which showed the relationship between the voltage VL4 which is an electrical potential difference, the 2nd-order differential value of voltage VL4, and the time after injection pulse OFF. 19, 20, 21, and 22, the same symbols are used for parts equivalent to those in FIGS. 1 and 2. 21 and 22, the same symbols are used for the same components as in FIG. In FIG. 23, the same symbols are used for parts equivalent to those in FIG.

  First, the structure and configuration of the drive unit of the fuel injection device in the closed state in which the valve body and the valve seat 118 are in contact with each other will be described with reference to FIGS. 19 and 20. From FIG. 19, the second valve body 1907 is provided with a first restricting portion 1910 at the top, and the second restricting portion 1908 is coupled to the second valve body 1907. A first member 1903 for supporting the initial position spring 1909 is joined to the second mover 1902 at the joint 1904 at the second mover 1902. The second movable element 1902 can be relatively moved between the first restricting portion 1910 and the second restricting portion 1908. In the closed state in which the second valve body 1907 and the valve seat 118 are in contact with each other, the second valve body 1907 includes a load by the spring 110 and a seat diameter at a contact position between the second valve body 1907 and the valve seat 118. A fluid force (hereinafter referred to as differential pressure) that is the product of the area of d2 and the fuel pressure acts in the valve closing direction. Further, the second movable element 1902 is urged in the valve closing direction by the load of the initial value spring 1909 and is stationary in contact with the second restricting portion 1908. In this valve-closed state, between the second restricting portion 1910 and the second movable element 1902. There is a gap 1901. Further, when the second valve body 1907 is in contact with the valve seat 118, there is no pressure difference between the upper part and the lower part of the second mover, and therefore no differential pressure acts on the second mover. Further, a vertical hole fuel passage 1905 is formed at the center of the second valve body 1907, and the fuel flows through the horizontal hole fuel passage 1906 to the downstream.

  The configuration of the drive device in the second embodiment will be described with reference to FIG. The difference between the driving device of the second embodiment and the driving device of the first embodiment is that the voltage measurement point for detecting the valve closing completion timing is changed from the voltage VL1 to the voltage VL, and the active low-pass filter 860 and the fuel are changed. By providing a capacitor C83 between the ground potential (GND) side terminal 2301 of the injection device 840 and the resistor R81, an analog differentiating circuit 2203 including capacitors C81 and C83, resistors R81 and R82, and an operational amplifier 820 is provided. The first-order differential processing of the voltage VL is performed in an analog manner by the driving device, and a signal of the first-order differential value of VL is input to the A / D conversion port of the CPU 801. In the analog differentiating circuit 2203, in the configuration in which the VL voltage is not divided, the potential difference between the ground potential (GND) side terminal of the solenoid 105 and the ground potential (GND) is detected. The maximum value is a high voltage value, for example, 60 V under the condition that a negative voltage is applied to the solenoid 105. Since the capacitor C1 is arranged between the measurement terminal 2301 for detecting the voltage VL and the operational amplifier 820, the voltage input to the operational amplifier 820 can be reduced, so that the operational amplifier 820 and the A / D converter of the CPU 801 are used. The required withstand voltage can be reduced, and the cost of the operational amplifier 820 and the CPU 801 can be reduced. Further, according to this configuration, it is possible to eliminate the resistor 853 necessary for dividing the voltage VL used in the first embodiment, which leads to cost reduction of the driving device. Further, by performing differentiation processing with the analog differentiation circuit 2203, high-frequency noise superimposed on the VL voltage of the driving device can be reduced, and the voltage value after the first-order differentiation processing is input to the CPU 801. The time resolution required for the A / D conversion port of the CPU 801 can be reduced, and the load of the filtering process and the digital differential operation process of the CPU 801 can be reduced.

  Further, the voltage VL2 for detecting the valve opening start timing and the valve opening completion timing is passed through the active low-pass filter 861, and the potential difference between the terminal 843 from which the high frequency noise component is removed and the ground potential 815 is expressed as the voltage VL3. Called. By inputting the voltage VL3 to the A / D conversion port of the CPU 801, a value obtained by dividing the voltage VL3 by the resistance value of the resistor 808 according to Ohm's law becomes a current flowing to the solenoid 105. Therefore, the current flowing to the solenoid 105 by the CPU 801 Can be detected. Further, according to the method of this embodiment, it is sufficient that the change in the slope of the current flowing through the solenoid 105, that is, the value of the current differential value can be detected by the driving device, so that the voltage VVL3 is differentiated to start and open the valve. Completion timing can be detected.

  Next, the valve opening operation of the fuel injection device 2305 in the second embodiment will be described with reference to FIGS. 19, 20, and 21. When current is supplied to the solenoid 105 and the magnetic attractive force acting on the second mover 1902 exceeds the load of the initial position spring 1909, the second mover 1902 moves in the valve opening direction, and the gap 1901 is 0. The second movable element 1902 collides with the second valve body 1907 at the timing when the second valve element 1907 is separated from the valve seat 118. As the second movable element 1902 moves in the valve opening direction, a shear resistance is generated between the outer diameter of the second movable element 1902 and the nozzle holder 101 in the second movable element 1902, and the second movable element 1902. A shear resistance acts on the child 1902 in the valve closing direction. However, the shear resistance can be reduced by increasing the gap between the outer diameter of the second movable element 1902 and the nozzle holder 101. Further, since the shear resistance force acting on the second mover 1902 is smaller than the magnetic attractive force that is the force in the valve opening direction, the second mover 1902 energizes the switching elements 805 and 808. The second movable element 1902 is accelerated in the valve opening direction by the magnetic attractive force generated when the boosted voltage VH is applied to the solenoid 105 and the current is supplied to the solenoid. Thereafter, the switching elements 805 and 806 are de-energized, and the boosted voltage VH in the negative direction is applied to the voltage Vinj between the terminals of the solenoid 105 to rapidly reduce the current flowing through the solenoid. Thereafter, the switching elements 807 and 806 are energized, the battery voltage VB is applied to the solenoid 105, and the second movable element 11902 is moved to the second valve body 1907 while the switching elements 807 and 806 are energized. The second valve body 1907 is started to open. The switching element 807, 806 is energized for a certain period of time after the second valve element 1907 starts opening or until the current value flowing through the solenoid 105 reaches a predetermined current value, whereby the second-order differential value of the current is maximized. The valve opening start timing can be detected as a value. In addition, as compared with the method in the first embodiment, the load by the spring 110 acts on the second valve body 1907 instead of the mover 102, so that the second valve body 1907 is started at the valve opening start timing. The change in acceleration of the movable element 1902 is large, and the change in the current gradient for detecting the valve opening start timing is large. This change in the slope of the current also occurs in the voltage VL2 for detecting the current flowing through the solenoid 105. Therefore, it is easy to detect the maximum value or the minimum value of the voltage VL2 after second-order differentiation of the voltage VL2, and as a result The detection accuracy of the valve opening start timing can be increased.

  Next, with reference to FIGS. 19, 20, and 21, an explanation of the operation of the second movable element 1902 and the second valve body 1907 when the valve body 114 in the second embodiment opens from the closed state and A method for detecting the valve opening completion timing will be described.

  In the state where the second valve body 1907 is in contact with the valve seat 118, the differential pressure does not act on the second mover 1902. Therefore, when current is supplied to the solenoid 105, the second mover 1907 is accelerated. After performing the operation and colliding with the second valve body 1907, the target lift is reached in a short time, and the second mover 1902 collides with the fixed core 107 at timing t2503. Unlike the fuel injection device 840 in the first embodiment, in the fuel injection device 2305 in the second embodiment, the load by the initial value spring 1909 acting on the second mover 1902 works in the valve closing direction. The second mover 1902 bounces as a result of the second mover 1902 colliding with the fixed core 107 after the valve body 1907 reaches the target lift multiple times such as 2506, 2507, 2508, It takes a long time for the bounce of the second mover 1902 to converge. As a result, inflection points due to the second movable element 1902 colliding with the fixed core 107 occur at timings t2502, t2503, and t2504 in the voltage VL3 for detecting the valve opening completion timing. There may be a case where a plurality of peaks in which the second order differential value protrudes in the positive direction, such as 2501, 2502, and 2503 (hereinafter referred to as peak 2501, peak 2502, and peak 2503). Even in this case, the valve opening completion timing can be detected by detecting the timing t2502 at which the second-order differential value of the voltage VL3 is maximized by the driving device for each fuel injection device of each cylinder. Also, the timing t2502 that triggers the acquisition period 2505 of the voltage VL3 for detecting the valve opening completion timing is set using the energization timing of the injection pulse or the energization / non-energization timing of the switching elements 805, 806, and 807. However, it is preferable that the above-described operation is performed after a certain period 2504 has elapsed since the energization / de-energization is performed. In particular, since the injection pulse output from the CPU 801 is generated inside the CPU 801, it can be easily used as a trigger for determining the period 2504. The acquisition period 2505 has a period in which individual variation in the valve opening completion timing of the fuel injection device of each cylinder can be detected, and in order to reduce the number of data points of the voltage VL3 input to the CPU 801, the acquisition period 2505 is previously set in the period 2504 and the acquisition period 2505. A set value may be set in the drive device. Further, when the fuel pressure supplied to the fuel injection device 2305 is changed, the differential pressure acting on the second valve body 1907 is changed, so that the valve opening completion timing is also changed. Accordingly, the period 2504 and the acquisition period 2505 are determined based on the target fuel pressure set by the CPU 801 of the driving device or the value detected by the driving device based on the output signal of the pressure sensor installed in the pipe upstream of the fuel injection device 2305. Good. As a result, even when the operating conditions change, the valve opening completion timing can be detected with high accuracy, and the data point sequence for taking in the voltage VL3 necessary for the detection into the CPU 801 can be reduced. Can be reduced. In addition, in the acquisition period 2505, there are a plurality of peaks in which the second-order differential value of the voltage VL3 protrudes in the positive direction, and the values of the second and third peaks 2502 and 2503 are larger than the value of the first peak 2501. If it is larger, the first peak 2501 may be stored in the drive device as the valve opening completion timing. By adopting such a configuration, it is possible to suppress erroneous detection of the valve opening completion timing while securing the acquisition period 2505 necessary for detecting the individual variation of the valve opening completion timing of the fuel injection device 2305 of each cylinder. Therefore, the detection accuracy of the valve opening completion timing and the correction accuracy of the injection amount can be improved. Further, as shown in FIG. 21, in a state where the second movable element 1902 comes into contact with the fixed core and is stationary, the second movable element 1902 is positioned between the lower end surface of the second movable element 1902 and the second restricting portion 1908. A gap 2101 is provided.

  Next, the second mover 1902 and the second mover 1902 when the second valve body 1907 in the second embodiment is closed from the state in which the displacement amount of the intermediate lift is maximized with reference to FIGS. 20, 22, and 24. The operation of the second valve body 1907 and the detection method of the valve closing completion timing will be described. 22 and 24, when the second valve body 1907 is closed from the open state, the load by the spring 110 and the differential pressure due to the fuel flow are applied to the second valve body 1907 as the force in the valve closing direction. The second movable element 1907 receives a force in the valve closing direction via the second valve body 1907, and the load of the initial position spring 1909 acts on the second movable element 1902 in the valve closing direction. ing. When the injection pulse is stopped, the switching elements 805 and 806 are de-energized, the negative boost voltage VH is applied to the solenoid 105, and the current flowing through the solenoid 105 is reduced, it acts on the second mover 1902. The magnetic attraction force decreases as the eddy current generated inside the magnetic circuit disappears. When the magnetic attractive force, which is the force in the valve opening direction acting on the second movable element 1902, is less than the force in the valve closing direction acting on the second valve element 1902 and the second movable element 1907, the second movable element The child 1902 and the second valve body 1907 start the valve closing operation. At the valve opening completion timing t2602 when the second valve body 1907 comes into contact with the valve seat 118, the second movable element 1902 moves away from the second valve body 1907 and continues to move in the valve closing direction. After that, the second movable element 1902 has a third interval between the lower end surface 2202 of the second movable element and the end surface of the second restricting portion 1908 at the moment when the second valve body 1907 and the valve seat 118 contact each other. At timing t2604 when the gap 2201 becomes 0, the second movable element 1902 collides with the second restricting portion 1908 at timing t2604 and stops. In this embodiment, timing t2601 when the injection pulse Ti is turned OFF is used as a trigger for taking in the voltage VL4 by the CPU 801, and data acquisition of the voltage VL4 is started after a predetermined time 2606 has elapsed since the injection pulse Ti was turned OFF. The voltage VL4 corresponding to the first-order differential value of the voltage VL may be input to the A / D conversion port of the CPU 801 only during 2607. Thereafter, the voltage VL4 captured by the CPU 801 is subjected to digital differentiation processing to calculate a first-order differential value of the voltage VL4. At this time, the first-order differential value of the voltage VL4 corresponds to the second-order differential value of the voltage VL.

  By detecting the first-order differential value of the voltage VL4 (corresponding to the second-order differential value of the voltage VL) by the driving device, the second valve body 1907 comes into contact with the valve seat 118, and the second movable element 1902 is The valve closing direction force acting on the second movable element 1902 that has been acting through the second valve element q907 so far at the timing of completing the valve closing at the moment of separation from the valve element 1907 of the second movable element 1902 Therefore, the acceleration of the second mover 1902 changes, and a first peak 2608 in which the first-order differential value of the voltage VL4 is negative is generated. Thereafter, at the moment when the second mover 1902 collides with the second restricting portion 1908, the second mover 1902 receives a reaction force due to contact with the second restricting portion 1908, and the acceleration changes greatly. As a result, a second peak 2609 having a negative first-order differential value of the voltage VL4 is generated. The value of the first-order differential value of the voltage VL4 of the first peak 2608 and the second peak 2609 depends on the clearance 1901 and the magnetic circuit shape, and the valve closing completion varies depending on the differential pressure due to the spring load and fuel pressure. This greatly depends on the speed of the second movable element 1902 at the timing. When the speed at the valve closing completion timing is small, the kinetic energy of the second mover 1902 at the valve closing completion timing becomes small, so the time from the valve closing completion timing until the second mover 1902 stops. In some cases, the value of the second peak 2609 is smaller than the value of the first peak 2608 in the first-order differential value of the voltage VL4. As described above, when searching for the minimum value of the first-order differential value of the voltage VL4 in the period 2607, one of the first peak 2608 and the second peak 2609 is detected. In such a case, the period 2607 is divided into a first period 2608 and a second period 2609, and the second valve body 1907 determines the minimum value of the first-order differential value of the voltage VL 4 in the first period 2608. It is determined as the valve closing completion timing in contact with the seat 118, and the minimum value of the first-order differential value of the voltage VL4 in the second period is set as the second regulating portion 1908 of the second valve element 1907 by the second movable element 1902. By detecting and determining each of the fuel injectors for each cylinder as the mover stationary timing that contacts the valve, it is possible to accurately detect the valve closing completion timing. During the valve closing operation, the second movable element 1902 continues to move in the valve closing direction until the second movable element 1902 collides with the second restricting portion 1908 after the second valve body 1907 comes into contact with the valve seat 118. When the next second injection pulse Ti for split injection is supplied while the second mover is moving in the valve closing direction, the previous injection pulse (referred to as the first injection pulse) Even if the equivalent second injection pulse is supplied, the second mover 1902 and the kinetic energy of the second mover 1902 change at the timing when the second injection pulse is supplied. The injection amount at the time of supplying the injection pulse Ti changes as compared with the case of supplying the first injection pulse width Ti. Therefore, it is preferable to detect the timing t2604 when the fuel injection device 2305 of each cylinder detected by the driving device stops and control the supply timing of the second injection pulse Ti. Further, the supply timing of the second injection pulse Ti may be adjusted in accordance with the individual fuel injection device 2305 having the longest timing t2604. According to the present embodiment, it is possible to reduce the interval between the first injection pulse and the second injection pulse under the condition of split injection in which fuel injection is performed a plurality of times during one intake / exhaust stroke. Since the injection amount of the first injection pulse and the second injection pulse can be accurately controlled, it is effective when the number of divided injections required is large. Further, the trigger for taking in the voltage VL4 may use the timing when the injection pulse Ti is turned on or the timing of energization / non-energization of the switching elements 805, 806 and 807.

  The fuel injection device 2305 and the driving device in the second embodiment may be used in combination with the fuel injection device 840 and the driving device in the first embodiment.

  A control method for correcting the injection amount of the fuel injection device 840 and the fuel injection device 2305 of the first and second embodiments according to the third embodiment of the present invention will be described with reference to FIGS.

  FIG. 25 shows a case where the valve body 114 or the second valve body 1907 is used while being held at the target lift position for a certain period of time when the fuel injection device 840 or the fuel injection device 2305 is driven by the method of the third embodiment. Terminal voltage Vinj of the fuel injection device 840 or the fuel injection device 2305, the drive current, the magnetic attractive force acting on the mover 102 or the second mover 1902, the valve acting on the valve body 114 or the second valve body 1907 FIG. 10 is a diagram showing a relationship between body driving force, displacement amount of the valve body 114 or the second valve body 1907, displacement amount of the movable element 102 or the second movable element 1902 and time. In the figure of the valve body driving force, the driving force in the valve opening direction is shown in the positive direction, and the driving force in the valve closing direction is shown in the negative direction. Further, in the driving current in the figure, a conventional current waveform that is generally used is indicated by a one-dot chain line. FIG. 26 shows the inter-terminal voltage Vinj, the drive current, the mover 102 or the second state in the operation state when the minimum injection amount is performed while the valve body 114 or the second valve body 1907 reaches the target lift. Magnetic attractive force acting on the movable element 1902, valve body driving force acting on the valve element 114 or the second valve element 1907, displacement amount of the valve element 114 or the second valve element 1907, the movable element 102 or the second movable element It is a figure showing the relation between the amount of displacement of child 1902 and time. In the figure of the valve body driving force, the driving force in the valve opening direction is shown in the positive direction, and the driving force in the valve closing direction is shown in the negative direction. 27 acts on the terminal voltage Vinj, the drive current, the mover 102 or the second mover 1902 when operating with an intermediate lift that realizes an injection amount smaller than the injection amount by the operation shown in FIG. Magnetic attraction force, valve body driving force acting on the valve body 114 or the second valve body 1907, displacement amount of the valve body 114 or the second valve body 1907, displacement amount of the movable element 102 or the second movable element 1902 It is the figure which showed the relationship of time. In the figure of the valve body driving force, the driving force in the valve opening direction is shown in the positive direction, and the driving force in the valve closing direction is shown in the negative direction. FIG. 28 is a diagram showing the relationship between the injection pulse width Ti and the fuel injection amount q when the current waveforms of the control method of FIGS. 25 to 27 are used.

  First, the operation when the valve body 114 or the second valve body 1907 is used while being held at the target lift position will be described with reference to FIG. From FIG. 25, when the injection pulse width Ti is supplied at time t2901, the switching elements 805 and 806 are energized, and the valve opening signal is turned ON, the boosted voltage VH is applied to the solenoid 105. Along with this, the current flowing through the solenoid 105 gradually increases, and the magnetic attraction force acting on the movable element 102 or the second movable element 1902 after the elapse of a certain delay time as the eddy current generated in the magnetic circuit disappears. Will increase. When the magnetic attraction force exceeds the valve closing force acting on the movable element 102 or the second movable element 1902, the movable element 102 or the second movable element 1902 starts to move, and the movement is gradually accelerated. However, in the fuel injection device 2305 in the second embodiment, the load by the set spring 110 acts on the second valve body 1907 in the closed state, and the second movable element 1902 is closed by the load by the initial position spring 1909. It is pushed in the valve direction. Next, at time 2902 when the current flowing through the solenoid 105 reaches the peak current value Ipeak, the switching elements 805 and 806 are deenergized to stop the application of the boosted voltage VH, and at the same time, the boosted voltage in the negative direction VH is applied. As a trigger for this operation performed at timing t2902, in addition to using the fact that the peak current value Ipeak has been reached as described above, a method of determining the boost voltage application time Tp in advance, and the peak current Ipeak reaching the peak current value Ipeak. Then, there is a method of setting after a certain time has elapsed. Depending on the circuit configuration, the boost voltage VH may fluctuate, and the resistance value, wiring resistance, inductance, etc. of the solenoid 105 of the fuel injection device 840 or the fuel injection device 2305 may vary, so the boost voltage application time Tp is fixed. In this case, the peak current value Ipeak varies. In order to give a stable valve opening force during the valve opening operation in consideration of the variation in the valve operation of the fuel injection device 840 or the fuel injection device 2305 of each cylinder, the control method of fixing the peak current value Ipeak Is good. On the other hand, a method of fixing the application time Tp is preferable in order to reduce variation in time for applying the valve opening force. Further, in the method of stopping the application of the boosted voltage VH after a certain time has elapsed since reaching the peak current value Ipeak, the effect of setting the peak current value Ipeak is obtained, and the current does not depend on the setting resolution of the peak current Ipeak. Since the cutoff time can be controlled, the current value can be adjusted more precisely, and the injection amount correction accuracy can be improved.

  Further, at the timing t2702 when the movable element 102 or the movable element 1907 collides with the valve element 114 or the second valve element 1907, the movable element 102 or the second movable element 1907 collides with the valve element 114 or the second valve element 1907. By doing so, the valve element 114 or the second valve element 1907 receives the kinetic energy of the mover 102 or the second mover 1907 and the impulse produced by the collision of the mover with the valve element. The second valve body 1907 performs a valve opening operation. At this time, energy input to the solenoid 105 during the period 1701 is converted into kinetic energy of the mover 102 or the second mover 1907. Thereafter, the valve element 114 or the second valve element 1907 reaches the target lift by the magnetic attractive force acting on the movable element 102 or the second movable element 1902, but the valve element 114 or the second valve element 1907 is displaced. A differential pressure (fluid force) according to the position acts in the valve closing direction. When the valve body 114 or the second valve body 1907 reaches the target lift position, a reaction force may occur due to the movable element 102 or the second movable element 1902 colliding with the fixed core 107. In order to reach the target lift with a holding current value lIh lower than the peak current value Ipeak while suppressing the valve opening speed of the valve body 114 or the second valve body 1907 in the boost voltage cutoff period T2, the reaction force is small, The mover 102 or the second mover 1902 does not bounce between the fixed core 107. Further, according to the configuration of the fuel injection device 840, the load of the return spring 112 acts in the valve opening direction that suppresses the bounce of the mover 102, and thus may occur when the mover 102 collides with the fixed core 107. There is an effect that the bounce of the movable element 102 can be suppressed.

  Further, after time t2702, when the current reaches 0 A while the negative boosted voltage VH is applied to the solenoid 105, the change in induced electromotive force generated by the change in current decreases, but at that time, the magnetic If the magnetic flux remains in the circuit, the magnetic attractive force and the disappearance of the magnetic flux are continued, and the voltage generated by the induced electromotive force is applied to the solenoid 105 as a negative voltage 2710. At the same time as the current flowing through the solenoid 105 decreases, the magnetic attractive force acting on the movable element 102 or the second movable element 1902 decreases, and the kinetic energy of the valve body 114 or the second valve body 1907 decreases. When the holding current value lh is supplied, the magnetic attractive force starts to increase again, and the valve body 114 or the second valve body 1907 reaches the target lift position.

  Also, once the peak current value Ipeak is reached, the current is rapidly cut off and lowered to the holding current value lh or less (referred to as a cut-off waveform), so that the valve body 114 or the second valve body 1907 reaches the target lift. The magnetic attraction force at the time of arrival can be made smaller than in the case of a current waveform (referred to as a conventional waveform) that shifts from the conventional peak current value Ipeak described in the drive current of FIG. 25 to the holding current value lh. Further, since the collision speed between the valve body 114 or the second valve body 1907 and the fixed core 107 can be reduced by reducing the magnetic attractive force, as shown in FIG. Compared to the above, the non-linearity generated in the injection amount characteristic can be improved, and the region where the relationship between the injection pulse width Ti and the fuel injection amount q is linear can be expanded in the direction in which the injection amount is small. When the valve body 1907 reaches the target lift, the controllable minimum injection amount can be reduced from the conventional waveform minimum injection amount 3002 to the cutoff waveform minimum injection amount 3003.

  Further, a valve opening delay time which is a time from the supply of the injection pulse Ti stored for each fuel injection device of each cylinder to the valve opening completion timing at which the valve body 114 or the second valve body 1907 reaches the target lift is set. It is preferable to adjust the peak current value Ipeak or the boost voltage application time Tp and the voltage cutoff time T2 for each fuel injection device of each cylinder. For example. For an individual whose valve opening delay time is early, since the valve opening speed is large, the boost voltage application time Tp is set short, and the time for the mover 102 or the second mover 1902 to start decelerating may be set earlier. . On the other hand, for an individual with a slow valve opening delay time, the boosted voltage application time Tp may be set longer to delay the time for the mover 102 or the second mover 1902 to start deceleration.

  When the current interruption waveform is used, when the injection pulse width Ti is OFF during the boost voltage interruption time Tp, the same current waveform is generated regardless of the magnitude of the injection pulse width Ti. Since a period of time is supplied to the solenoid 105 of 2305, a dead zone Tn is generated in which the fuel injection amount q does not change even if the injection pulse width Ti is increased. In the injection amount characteristic of the cutoff waveform shown in FIG. 28, an intermediate lift region Tharf where the valve body 114 does not reach the target lift, and an injection pulse width Ti region after 3003 where the valve body 114 reaches the target lift and is driven However, since the non-linearity of the injection amount characteristic that has occurred in the injection amount characteristic of the conventional waveform is improved, the relationship between the injection pulse width and the fuel injection amount q is different. Always has a positive relationship, and the fuel injection amount q increases as the injection pulse width increases. In order to simplify the injection amount control algorithm mounted on the CPU 801 of the driving device, it is necessary to continuously increase the injection amount as the engine speed or the engine load increases. The fuel injection amount q needs to increase as the injection pulse width Ti increases. In such an engine, by using the control method in the third embodiment, it is possible to appropriately control the fuel injection amount q required as the engine speed or the engine load increases, and it is easy to control the injection amount. It becomes. In the case of using the conventional waveform, the deviation value between the ideal straight line 3001 and the fuel injection amount q obtained from the injection amount in the region where the relationship between the injection pulse width and the injection amount is substantially linear varies in the positive and negative directions. In the region where the injection amount characteristic is non-linear, the relationship between each injection pulse width Ti and the fuel injection amount q needs to be grasped by the driving device, so that the valve closing completion timing is detected for each injection pulse width Ti. It is necessary to memorize | store in a drive device by the fuel-injection apparatus of each cylinder as valve closing delay time. On the other hand, in the control method using the cutoff waveform in the third embodiment, the relationship between the injection pulse width Ti and the fuel injection amount q has a positive correlation in the intermediate lift region Tharf and the region after reaching the target lift. Therefore, detection information of the valve closing completion timing of two points in each of the intermediate lift region Tharf and the region reaching the target lift, and detection information of the valve opening completion timing and valve opening start timing of one point in the region reaching the target lift Based on this, it is possible to calculate a deviation value from the required injection amount, and to reduce the memory capacity necessary for storing the calculation load and individual information of the CPU 801 or IC 802 necessary for detecting the valve operation It is possible to simplify the algorithm for correcting individual variations in the injection amount given to the CPU 801 or the IC 802. In addition, when there is a request for an injection amount smaller than the controllable minimum injection amount 3003 under the condition that the valve body 114 or the second valve body 1907 reaches the target lift, the injection pulse width is smaller than the period of the dead zone Tn. The dead zone Tn may be set in advance for each fuel injection device 840 or fuel injection device 2305 of each cylinder so that Ti is used.

  Specifically, when adjusting the peak current value Ipeak or the boost voltage application time Tp and the voltage cut-off time T2, the parameter is adjusted in a feedback manner by storing the valve opening delay time Ta of each cylinder in the drive device. Thus, it becomes possible to cope with individual variations in the operating characteristics of the fuel injection device 840 or the fuel injection device 2305, changes due to deterioration, and the like, and stable operation can be realized. In the fuel injection device 840 or the fuel injection device 2305, the valve opening completion timing varies due to the influence of variation in dimensional tolerance. When the same shut-off waveform is supplied to the solenoid 105 for an individual whose valve opening completion timing is late and an early one, in the individual whose valve opening completion timing is early, the boost voltage cutoff timing t2702 which is the timing when the peak current value Ipeak is shut off Even if the current is cut off, the deceleration of the movable element 102 or the second movable element 1902 is not in time, the collision speed between the movable element 102 or the second movable element 1902 and the fixed core 107 increases, and the injection amount characteristic is Non-linearity may occur. In addition, in an individual whose valve opening completion timing is late, when the switching elements 805 and 806 are de-energized at the end timing of the boost voltage cutoff time Tp and the current flowing through the solenoid 105 is reduced, the valve body 114 or the second valve body The magnetic attractive force acting on the mover 102 or the second mover 1902 necessary for the 1907 to reach the target lift cannot be secured, and the valve body 114 or the valve body 1907 does not reach the target lift position. Therefore, using the information of the valve opening delay time stored in the drive device, after the valve body 114 or the second valve body 1907 starts to open for each fuel injection device 840 or fuel injection device 2305 of each cylinder, When a certain amount of displacement is reached, the switching elements 805 and 806 are de-energized, the negative boost voltage VH is applied to the solenoid 105, and the timing at which deceleration starts from the timing of valve opening completion is equalized. In addition, the boosted voltage application time Tp and the voltage cutoff time T2 may be adjusted. In addition, the peak current value Ipeak automatically changes by changing the boost voltage application time Tp, but the setting of the peak current value Ipeak is changed for each fuel injector 840 or fuel injector 2305 to increase the boost voltage. The marking time Tp may be adjusted. By adjusting the peak current value Ipeak for each individual, as compared with the case where the boosted voltage application time Tp is adjusted, the current flowing through the solenoid 105 and the valve caused by the voltage value of the boosted voltage VH of the driving device vary. Since the variation in operation can be minimized, an appropriate deceleration timing can be adjusted for each fuel injection device 840 or fuel injection device 2305 of each cylinder. By appropriately correcting the peak current value Ipeak and the drive voltage cutoff time T2 for each fuel injection device of each cylinder, individual variations in speed when the movable element 102 or the second movable element 1902 and the fixed core 107 collide with each other. Therefore, it is possible to reduce the drive sound when the valve is opened due to a collision, and the engine can be silenced. Further, by reducing the collision speed between the movable element 102 or the second movable element 1902 and the fixed core 107, the impact force acting on the collision surface between the movable element 102 or the second movable element 1902 and the fixed core 107 is reduced. Since it is possible to prevent deformation and wear of the collision surface, changes in the target lift amount due to deterioration can be suppressed. Further, according to the effect of the present embodiment, the collision speed between the movable element 102 or the second movable element 1902 and the fixed core 107 can be reduced and kept constant regardless of the individual fuel injection devices of the respective cylinders. Therefore, the hardness of the material necessary to prevent the deformation and wear of the collision surface can be reduced, and the end surface of the movable element 102 or the second movable element 1902 on the fixed core 107 side or the movable element of the fixed core 107 Since the plating process formed on the end surface on the 102 side is not necessary, a significant cost reduction can be achieved. By not performing the plating process, the flow rate per unit time due to the individual variation of the target lift caused by the variation of the plating thickness and the fluid gap between the movable element 102 and the fixed core 107 in the valve open state Since the variation of the squeeze force that accompanies this variation can be suppressed, the accuracy of the injection amount can be increased.

  Further, when the valve body 114 or the second valve body 1907 reaches the target lift, the movable element 102 or the second movable element 1902 and the fixed core 107 come into contact with each other, and the valve body 114 or the second valve body 1907 is moved. When stationary at the target lift position, the fuel injected from the fuel injection device 840 or the fuel injection device 2305 has a constant flow rate, and the injection amount can be increased in proportion to the increase in the injection pulse width Ti, so that the injection amount can be accurately controlled. It becomes possible to do.

  Further, the current cut-off waveform is obtained by correcting either the peak current value Ipeak or the boost voltage application time Tp and the voltage cut-off time T2 so that the injection amounts are equal in the fuel injection devices of the respective cylinders. The value of the dead zone Tn of the injection amount characteristic generated when it is used differs for each fuel injection device of each cylinder. When one value of the peak current value Ipeak or the boosted voltage application time Tp and the voltage cutoff time T2 are determined using the detection information, the dead zone Tn is determined. Therefore, by configuring the CPU 801 or the IC 802 so that the dead zone Tn can be set to a different value for each fuel injection device 840 or fuel injection device 2305 of each cylinder, the injection pulse width Ti is small and the valve body 114 reaches the target lift. Since it is possible to control by continuously changing from the intermediate lift region Tharf to the injection amount after the minimum injection amount 3003 after the valve body reaches the target lift, it is possible to control the injection amount according to the engine operating conditions. It can be carried out.

  The valve closing operation is performed by de-energizing the switching elements 807 and 806 at time t2704 when the injection pulse width Ti that is the valve opening signal time is stopped, so that the boosted voltage VH in the negative direction is applied to the solenoid 105, The current flowing through the solenoid 105 is rapidly reduced, and the magnetic attractive force is reduced. The operation of the valve body 114 or the second valve body 1907 in the valve closing direction is started at time t2705 when the magnetic attractive force is less than the force in the valve closing direction, and the valve closing is completed at time t2706. However, in the fuel injection device 2305, after the second valve body 1907 is closed, the load by the set spring 110 continues to act in the valve closing direction of the valve body driving force of the second valve body 1907. The force in the valve closing direction of the valve body driving force before the start of valve opening and after the completion of valve closing shown in FIG. 25 indicates the valve body driving force when the fuel injection device 2305 is used. In addition, the valve closing completion delay time Tb, which is the time from when the injection pulse width Ti is turned ON to when the valve body 114 or the second valve body 1907 is closed, is detected and stored by the driving device, and the target setting is performed. If there is a deviation from the delay time of the value, the setting of the holding current value lh at the target lift position may be increased or decreased to match the standard delay time. In addition, when correcting the individual variation of the valve closing completion delay time after correcting the drive current and driving voltage in the fuel injection device of each cylinder, the injection pulse width Ti is corrected and the valve closing completion delay time is large. The valve body 114 or the second valve body 1907 is actually opened by reducing the injection pulse width Ti and reducing the valve closing completion delay time by increasing the injection pulse width Ti. The actual injection period (Tb−Ta ′) can be controlled to an actual injection period necessary for realizing the required injection amount, and the correction accuracy of the injection amount can be improved.

  FIG. 26 shows an operation state when the minimum injection amount is performed while the valve body 114 or the second valve body 1907 reaches the target lift by the operation procedure of this method. At time t2801, the valve opening signal, that is, the injection pulse is turned ON, the switching elements 805 and 806 are energized, the boosted voltage VH is applied to the solenoid 105 from the second voltage source, and the movable element 102 or the second movable element 1902 is applied. Generate magnetic attraction. After that, when the peak current Ipeak is reached, or when the boosted voltage application time Tp is reached, the energization of the switching elements 805 and 806 is stopped to stop the application of the boosted voltage VH and the negative boosted voltage. By applying VH, the current flowing through the solenoid 105 is rapidly reduced, and the magnetic attractive force acting on the mover 102 or the second mover 1902 is lowered. After the set time of the voltage cutoff time T2 for cutting off the voltage in the driving direction, that is, the voltage in the positive direction, the switching elements 806 and 807 are energized, and the valve is opened at the timing when the voltage is applied from the battery voltage VB to the solenoid 105. When the injection pulse width Ti, which is the signal time, is turned ON, the valve body 114 or the second valve body 1907 that has reached the target lift position before and after that time has a magnetic attraction force that generates a force in the valve closing direction of the valve body driving force. After the lower timing, the operation moves to the valve closing direction, and the valve closing operation is performed without stopping at the target lift position. In order to perform the operation of the minimum injection amount at the full lift, when the injection pulse width Ti is increased in the operation at this time, the time during which the valve body 114 is stationary at the target lift position is long. Need to be. That is, ideally, at the minimum injection amount, the rest time at the target lift position is almost as close to 0 seconds, and when the valve opening signal time, that is, the injection pulse width Ti is increased, the increased time is The time during which the valve body is stationary at the target lift position is lengthened, and the valve closing completion timing is increased and the injection amount is increased in accordance with the increase of the stationary time, so that the injection pulse width Ti and the fuel injection amount q are linear. It is good to control so that it may become a general relationship.

  Further, when the fuel pressure supplied to the fuel injection device 840 or the fuel injection device 2305 changes, the peak current value Ipeak necessary for the valve body 114 or the second valve body 1907 to reach the target lift, and the valve body 114 Alternatively, the holding current value lh that can hold the second valve body 1907 in the open state changes. When the fuel pressure increases, in a state where the valve body 114 or the second valve body 1907 is closed, a force obtained by multiplying the pressure receiving area of the seat diameter and the fuel pressure acts on the valve body 114 or the second valve body 1907. Therefore, the kinetic energy of the mover 102 or the mover 1902 necessary for the valve element 114 or the second valve element 1907 to start opening changes. When the movable element 102 or the movable element 1907 collides with the valve element 114 or the second valve element 1907 and the displacement of the valve element 114 or the second valve element 1907 is started, the valve element 114 or the second valve element 1907 is started. The flow rate of the fuel flowing through the fuel seat portion of the valve body 1907 increases, and the pressure of the fuel flowing in the vicinity of the seat portion rapidly decreases due to the pressure drop (static pressure drop) based on Bernoulli's theorem. Alternatively, the pressure difference between the pipe side and the tip of the second valve body 1907 increases, and the differential pressure acting on the valve body 114 or the second valve body 1907 increases. The required peak current value Ipeak, voltage cutoff time T2, and holding current value lh may be adjusted according to the increase / decrease in the differential pressure. When the driving current holding current value lh is kept constant under a wide range of fuel pressure conditions with different engine loads, the valve element 114 or the second valve element 1907 can be held open at a high fuel pressure. Thus, it is necessary to set a high holding current value lh that can generate a magnetic attractive force acting on the movable element 102 or the second movable element 1902. When the valve body 114 or the second valve body 1907 is driven with a high holding current value lh at a low fuel pressure so as to reach the target lift, when the injection pulse width Ti is stopped, The magnetic attractive force generated in the second mover 1902 increases, the valve closing delay time increases, and the injection amount also increases. Therefore, as a configuration in which a command signal is sent from the ECU 120 to the drive circuit 121, the fuel pressure is adjusted using a signal from a pressure sensor attached to the fuel pipe upstream of the fuel injection device 840 or the fuel injection device 2305 detected by the ECU. Accordingly, an appropriate holding current value lh may be set.

  In addition, the individual variation of the fuel injection device 840 and the fuel injection device 2305 in each cylinder is similar to the change in the fuel pressure, and the valve body 114 or the second valve body 1907 is opened due to the variation in the load of the spring 110. The holding current value Ih necessary for holding also changes. In an individual having a large load due to the spring 110, the magnetic attraction force required to hold the valve body 114 or the second valve body 1907 in an open state is increased, and therefore it is necessary to set the holding current value Ih large. The load of the spring 110 is adjusted in the process of adjusting the injection amount of the fuel injection device 840 or the fuel injection device 2305. Therefore, since there is a strong correlation between the valve opening delay time, the valve closing delay time, and the load of the spring 110, the load of the spring 110 can be estimated from the opening / closing valve delay time. By storing information on the load due to the spring 110 estimated for each cylinder in the drive device, the timing for decelerating the mover 102 or the second mover 1907 based on the information on the load due to the spring 110 and the valve opening delay time. By correcting the peak current value Ipeak or boosted voltage application time Tp and voltage cutoff time T2 for each fuel injection device 840 or fuel injection device 2305 of each cylinder, the fixed core of the movable element 102 or the second movable element 1902 Therefore, the continuity of the injection amount characteristic from the intermediate lift to the full lift operation can be ensured, so that the injection amount can be easily controlled.

  In addition to the adjustment of the peak current value Ipeak, the boost voltage marking time Tp, and the voltage cutoff time T2 for reducing the individual variation of the fuel injection device 804 and the fuel injection device 2305 of each cylinder, the current waveform is adjusted by the fuel pressure. And effective. As the fuel pressure increases, the differential pressure due to the fuel pressure acting on the second valve body 1907 increases, so the timing at which the second valve body 1907 decelerates after the peak current value Ipeak is cut off is also increased. The bounce of the second valve body 1907 generated when the second movable element 1902 collides with the fixed core 107 after the valve body 1907 reaches the target lift position is also reduced. Therefore, by increasing the peak current value Ipeak according to the increase in the fuel pressure, the second movable element is secured while ensuring the peak current value Ipeak necessary for the second valve body 1907 to reach the target lift. The collision speed between 1902 and the fixed core 107 can also be reduced, the non-linearity of the injection quantity characteristic can be reduced, and the injection quantity variation can be reduced. Further, when the peak current value Ipeak is increased, the timing at which the switching elements 805 and 806 are de-energized, that is, the timing at which the application of the boost voltage VH is stopped is delayed, and the timing at which the voltage cutoff time T2 is started is also linked. Become slow. The voltage cut-off time T2 is preferably configured to become shorter as the fuel pressure increases. With such a configuration, when the differential pressure acting on the valve body 114 or the second valve body 1907 increases as the fuel pressure increases, the movable element 102 or the second movable element 1902 and the fixed core Since the collision speed 107 becomes smaller and the timing required for deceleration is also delayed, it is possible to set an appropriate deceleration timing. Since the fuel pressure and the differential pressure acting on the valve body 114 or the second valve body 1907 have a linear relationship, the peak current value Ipeak or the boosted voltage marking time Tp and the holding current value lh depend on the fuel pressure. It is preferable to give a correction coefficient for determining the value to the ECU or the drive circuit in advance. Further, the peak current value Ipeak and the holding current value lh described above are adjusted for each fuel injection device 840 or fuel injection device 2305 of each cylinder and for each fuel pressure supplied to the fuel injection device 840 or fuel injection device 2305. By doing so, since the current to be used can be reduced, the heat generation of the solenoid 105 of the fuel injection device 840 or the fuel injection device 2305 and the heat generation of the ECU can be reduced, and the energy consumption can be reduced. In addition, since the time during which the boost voltage VH is applied is reduced, the load on the boost circuit can be reduced, and the boost voltage VH at the time when the next injection pulse width is requested in the divided injection can be kept constant. Therefore, it is possible to accurately control the injection amount.

  Next, FIG. 27 shows an operation for using a region where the valve body 114 does not reach the target lift (referred to as an intermediate lift region) by the control method of the second embodiment. In this operation, in order to realize an injection amount that is smaller than the minimum injection amount when the target lift is reached, the peak current value Ipeak is lowered from the standard set value in accordance with the reduction in the injection amount. Reduce the injection amount. That is, when an injection amount smaller than the injection amount by the operation shown in FIG. 26 is realized, the injection pulse width Ti that is the valve opening signal time, the set value of the peak current value Ipeak that determines the time for applying the boost voltage, It is preferable to change the set value of the voltage application time Tp. As shown in FIG. 26, by setting the set value Ip ′ smaller than the standard peak current value Ipeak, the application of the boost voltage VH is stopped at time t2902 when the current flowing through the solenoid 105 reaches Ip ′. As a result, the boosted voltage VH in the negative direction is applied to the solenoid 105, and the current flowing through the solenoid 105 rapidly decreases, thereby reducing the magnetic attractive force. However, in a region where the fuel to be injected is small and the displacement amount of the valve body 114 is small, the movable element 102 or the second movable element 1902 collides with the valve body 114 or the second valve body 1907, so Since the valve body 114 or the second valve body 1907 starts to open due to the impulse and kinetic energy received by the second valve body 1907, the positive pressure to the solenoid 105 is not changed before time t2904 when the valve body 114 starts to open. The voltage application in the direction of is preferably stopped. This positive voltage stop occurs when the switching pulse 805 and switching element 806 are energized and the boosted voltage VH is applied to the solenoid 105 and the switching element 805 and switching element 806 are de-energized. The boosted voltage application time Tp until the boosted voltage VH in the negative direction is applied to the solenoid 105 may be controlled by the set value Ip ′. The kinetic energy generated in the movable element 102 at a timing before the valve body 114 starts to open can be controlled by the boost voltage application time Tp or the set value Ip ′, and the displacement amount of the valve body 114 can be controlled. It becomes possible. Further, in this intermediate lift operation, since the valve body 114 does not reach the target lift, the displacement amount of the valve body 114 is not defined by the mechanism, and individual variations in the injection amount are likely to occur due to slight changes in the fuel pressure and the like. Accordingly, the valve closing completion timing t2905, which is the time when the first-order differential value of the voltage VL4 becomes the minimum value or the time when the second-order differential value of the voltage VL becomes the minimum value after the injection pulse is turned on, is the fuel of each cylinder. By detecting each injection device and storing it in the drive device, the ECU 120 or EDU 121 checks whether or not the valve closing completion timing or injection period for realizing the required injection amount is coincident with the target value. If so, it is possible to increase the accuracy of the actual injection amount with respect to the required injection amount by adjusting the set value Ip ′ of the peak current to be increased or decreased during the next injection. Similarly, in the case of the method of setting the boost voltage application time Tp, the boost voltage is adjusted so that the valve closing completion timing t2904 is detected by the driving device and the valve closing completion timing or injection period for realizing the required injection amount is met. By adjusting the application time Tp, the accuracy of the actual injection amount with respect to the required injection amount can be increased.

  A control method for correcting the injection amount in the fourth embodiment according to the present invention will be described with reference to FIGS.

  FIG. 29 shows the valve opening start timing Ta ′ and valve closing of the valve body 114 or the second valve body 1907 under the condition that the same injection pulse width Ti is supplied to the individual fuel injection devices 1, 2, and 3 of each cylinder. As a result of correcting the injection pulse, the drive voltage, and the drive current so that the injection periods (Tb−Ta ′) match for individuals with different completion timings Tb, the drive voltage, drive current, and valve body displacement amount of each individual as a result of correction. It is the figure which showed the relationship between time. 29, the valve body displacement amounts of the individual 1 and the individual 3 when the same injection pulse width, drive voltage, and drive current as those of the individual 2 are supplied are described.

  As described with reference to FIG. 6 of the first embodiment, even if the same injection pulse width is supplied, the timing of the valve operation, that is, the valve body 114 or the second valve body for each fuel injection device of each cylinder due to the influence of fluctuations such as dimensional tolerance. The valve opening start timing Ta ′ and the valve closing completion timing Tb of the valve body 1907 are different, the valve body 1907 is separated from the valve seat 118, and the actual injection period (Tb−Ta ′) during which fuel is injected is individual for each individual. As a result, the individual variation in the injection amount occurs. In the control method in the third embodiment, the injection amount is detected by using the detection information stored in the drive device of the valve opening start timing, the valve opening completion timing, and the valve closing completion timing described in the first embodiment and the second embodiment. A fuel injection control method for suppressing the individual variation of the fuel injection will be described. With reference to FIG. 25, a method for correcting individual variations in the injection amount at the minimum injection amount with the smallest injection amount at a certain fuel pressure will be described. In the individual 1 (before correction) with the earlier valve opening start timing Ta ′, when the same injection pulse width, drive voltage, and drive current as those in the individual 2 are supplied, the valve body displacement amount at the timing when the current supply is stopped compared to the individual 2 Is large, the valve closing completion timing Tb is delayed. As a result, the injection period is longer than that of the individual 2, and the injection amount is also increased. Further, in the individual 1 (before correction) with the slow valve opening start timing Ta ′, when the same injection pulse width, drive voltage, and drive current as those in the individual 2 are supplied, the valve body displacement amount at the timing when the current supply is stopped is Therefore, the valve closing completion timing Tb is earlier, and as a result, the injection period is shorter than that of the individual 2, and the injection amount is also reduced. For the individual 1 (before correction) having a long injection period, the injection pulse Ti is shortened, the period during which the boost voltage VH is applied is shortened to Tp1, or the peak current value Ipeak of the drive current is set to Ip1 ′. Thus, it is preferable to correct the above parameters so as to coincide with the injection period 2702 of the individual 2. On the other hand, for the individual 3 having a short injection period (before correction), the injection pulse Ti is lengthened, the period for applying the boosted voltage VH is lengthened as Tp3, or the peak current value Ipeak of the drive current is set to It is preferable to correct the above parameters so as to match the injection period 2702 of the individual 2 by increasing it to Ip3 ′. When the injection period is corrected using the peak currents Ip1 ′, Ip2 ′, Ip3 ′ of the drive current, there is a change in resistance due to a temperature change of the solenoid 105 or a change in the voltage value of the boost voltage VH. However, the variation of the displacement amount of the valve body 114 or the second valve body 1907 can be suppressed to the minimum, and the unintended fluctuation of the injection period accompanying the environmental change can be suppressed. In addition, when the injection period is corrected using the application time Tp1, Tp2, and Tp3 of the boost voltage, the time resolution can be reduced as compared with the method using the peak current of the drive current, so that the injection period can be reduced. There is an effect of improving the correction accuracy. This is because the setting resolution of the peak current value depends on the resistance value of the resistor 808 or 812 for detecting the current value. As the resistance value is reduced, the peak current value setting resolution is improved. However, if the current value is too small, detection by the IC 802 becomes difficult. Further, the drive voltage stop timing for adjusting the injection period may be set so that a certain time elapses after the target current value is reached. Due to this effect, even if there is a change in the resistance of the solenoid 105, it is possible to suppress unintended fluctuations in the injection period and improve the time resolution of the drive voltage stop timing. It is possible to improve the correction accuracy of the injection period and the correction accuracy of individual variations in the injection amount.

  Next, a method for controlling the injection amount after adjusting the injection period with the minimum injection amount will be described with reference to FIGS. 30 and 31. FIG. 30 is a diagram illustrating a method of adjusting the injection amount after adjusting the injection period with the minimum injection amount. FIG. 31 is a diagram showing the relationship between the injection pulse and the injection amount after adjusting the injection period with the minimum injection amount.

  From FIG. 30, Tp at the minimum injection amount is adjusted for each fuel injection device 840 or fuel injection device 2305 of each cylinder so that the injection period matches as described above. Thereafter, in order to control the injection amount at the intermediate lift, after the T2 end timing t2804, the switching elements 805 and 806 are energized, the boosted voltage VH is applied to the solenoid 105, and the holding current lh is shifted. Thereafter, the energization time of the injection pulse Ti is increased, and the valve body 114 or the second valve body 1907 is reached to the target lift position in contact with the fixed core 107. Changes in valve closing completion timing by increasing the injection pulse Ti in the fuel injection device 840 or the fuel injection device 2305 of each cylinder after Ti2 and Ti3 performing intermediate lift operation after the injection pulse width Ti1 at the minimum injection amount When the amount varies from individual to individual, learning control is performed so that the holding current value lh2 is increased and the magnetic attractive force is increased for the individual with a small amount of change in the valve closing completion timing so that the injection period is met. On the other hand, for an individual with a large change amount of the valve closing completion timing, learning control may be performed so that the holding current value Ih1 is reduced to reduce the magnetic attractive force so that the injection period is matched. In this way, by adjusting the current value of the holding current lh for each cylinder, it is possible to stably reach the target lift, and it is possible to improve the injection amount correction accuracy.

  By controlling the displacement amount of the valve body 114 or the second valve body 1907 by the method described above, the injection amount characteristic shown in FIG. 31 is the same as the injection pulse width Ti in the section 3401 of the conventional waveform in the intermediate lift region. With respect to the gradient of the injection amount, the gradient between the injection pulse width Ti and the fuel injection amount in the section Tharf2 is reduced, and the intermediate lift region until reaching the target lift is expanded from Tharf1 to Tharf2. In the section 3401 with the intermediate lift of the conventional waveform, the injection amount changes greatly with respect to the change of the injection pulse width. Therefore, when performing the fine injection amount control, the injection pulse width Ti or the time of the boost voltage application time Tp The resolution must be set finely and a driving device with a high number of clocks of the CPU 801 must be used, leading to an increase in the cost of the driving device. In addition, since the fuel injection amount with respect to the injection pulse width Ti is nonlinear between the section 3401 with the intermediate lift and the target lift region, in order to control the injection amount, the injection period with the injection pulse width Ti at each point This is necessary to detect the information of the drive unit, which causes pressure on the storage capacity of the driving device, and furthermore, the injection amount after the end of the section 3401 may change greatly due to changes in environmental conditions, etc. It is difficult to improve accuracy and robustness. According to the control method in the third embodiment, the difference between the inclination of the injection pulse width Ti and the fuel injection amount q in the intermediate lift region and the inclination of the injection pulse width Ti and the fuel injection amount q after reaching the target lift are conventionally calculated. Compared to the control method using the waveform, the relationship between the injection pulse width Ti and the fuel injection amount q is linear after the target lift from the intermediate lift region, so that the injection amount is corrected and controlled. There is an easy advantage. As described above, as a result of individual adjustment of the drive voltage and current waveform by the fuel injection device 840 or the fuel injection device 2305 of each cylinder, the injection amount characteristic becomes a characteristic that is translated in the injection pulse width Ti direction. , And a shift 3401 for the parallel movement. However, since the injection period for determining the fuel injection amount q can be detected for each cylinder by the driving device, individual variations in the injection amount can be achieved by correcting the deviation 3401 for the parallel movement with the injection pulse width Ti for each cylinder. Can be corrected and controlled. Further, when the relationship between the injection pulse width and the fuel injection amount in the intermediate lift region is a linear approximation, if there are two pieces of information on the injection period for detecting the inclination, the inclination and intercept of the correction formula Can be derived. In the target lift region, the fuel injection amount q increases linearly as the injection pulse width Ti increases, so the relationship between the injection pulse width Ti and the fuel injection amount q is approximated by a linear approximation function. The slope and intercept of the function can be derived from information on two or more injection periods. Further, the injection pulse width Ti for switching from the intermediate lift to the target lift can be calculated as a point where the fuel injection amount q of the primary function at the intermediate lift and the primary function at the full lift overlap, and the injection in the intermediate lift region It is preferable that the correction formula for the amount and the correction formula for the injection amount after the target lift can be switched.

  The fifth embodiment of the present invention is an embodiment showing an example in which the fuel injection device described in Embodiments 1 to 4 and the control method thereof are mounted on an engine.

  FIG. 32 is a configuration diagram of an in-cylinder direct injection gasoline engine, and the fuel injection devices A01A to A01D are installed such that fuel spray from the injection holes is directly injected into the combustion chamber A02. The fuel is boosted by the fuel pump A03, sent to the fuel pipe A07, and delivered to the fuel injection device A01. The fuel pressure varies depending on the balance between the amount of fuel discharged by the fuel pump A03 and the amount of fuel injected into each combustion chamber by the fuel injection device provided to each cylinder of the engine, but based on information from the pressure sensor A04. The discharge amount from the fuel pump A03 is controlled with a predetermined pressure as a target value.

  The injection of fuel is controlled by the injection pulse width sent from the ECU engine control unit (ECU) A05, and this injection pulse is input to the drive circuit A06 of the fuel injection device, and the drive circuit A06 is based on a command from the ECU A05. Thus, the drive current waveform is determined, and the drive current waveform is supplied to the fuel injection device A01 for a time based on the injection pulse.

  Note that the drive circuit A06 may be mounted as a component or a board integrated with the ECU A05.

  The ECU A05 and the drive circuit A06 have the ability to change the drive current waveform depending on the fuel pressure and operating conditions.

  In such an engine, when the ECU A05 has the capability of detecting the opening and closing operations of the fuel injection device A01 as described in the first to seventh embodiments, the engine can be easily controlled, A method for reducing engine exhaust by reducing exhaust or reducing variation in combustion pressure between cylinders will be described.

  In ECUA05 used for the engine described in FIG. The injection pulse width of the fuel injection device A01 is corrected so that the amount of fuel injected from the fuel injection devices A01A to A01D approaches the value required by the ECU A05. That is, in a multi-cylinder engine, drive pulses having different widths corrected for each cylinder are given to the respective fuel injection devices.

  For example, for a fuel injection device that drives with a short pulse width for a fuel injection device that injects a large amount of fuel when the same command pulse width is given, and for a fuel injection device that injects a small amount of fuel when the same pulse width is given Drive with a long pulse width. By having an operation mode in which such correction is performed for each cylinder, it is possible to suppress variations in the fuel injection amount between the cylinders.

  Further, in the ECU A05 shown in FIG. 32, the drive current supplied to the fuel injection devices A01A to A01D of each cylinder is supplied as a waveform adjusted for each fuel injection device.

  Each current waveform is set so that the valve behavior of each fuel injection device A01A to A01D is reduced so that the rebound behavior at the time of valve opening is reduced. As a result, the relationship between the injection pulse width and the injection amount is a straight line. It can be set so that the range of the pulse width approaching is widened.

  For example, in order to reduce the rebound behavior at the time of valve opening, the time for supplying the boosted voltage VH from the boosted voltage source to the solenoid 105 in the drive waveform or the peak current value Ipeak is set to the energization / switching of the switching elements 805 806 807 By controlling the deenergization, adjustment is made in accordance with the valve opening timing of the fuel injection device of each cylinder, and the energization from the boosting power supply is cut off during the valve opening, and the valve is set to decelerate. For example, for a fuel injection device that opens early when a certain current waveform is applied, the timing for stopping the energization from the boost power supply is advanced, and for the fuel injection device 840 or fuel injection device 2305 that opens late, the pressure is increased. Set the timing to stop energization from the power supply late. In this way, by using a drive waveform that cuts off the energization from the boosting power source and decelerates the valve opening operation, it is possible to reduce the change in the injection amount with respect to the change in the injection pulse width Ti in the region of the minute injection amount. This also has the effect of facilitating correction of the injection amount based on the injection pulse width Ti.

  By providing the drive current waveform in which the valve element 114 decelerates in this manner in accordance with the variation in the valve opening completion timing of the cylinder fuel injection device 840 or 2305, a current waveform suitable for the fuel injection device of each cylinder can be provided. Thus, the range in which the relationship between the injection pulse and the injection amount is linear can be increased.

  Moreover, it is good to adjust the energization current value (holding current value) for hold | maintaining a valve opening state among drive waveforms according to the valve closing timing of each fuel injection apparatus. When the valve closing timing obtained when the fuel injection device is driven with a certain driving current waveform is late, the holding current value is set small, and when the valve closing timing is early, the holding current value is relatively large. Set. Thus, by setting the holding current value in the drive current waveform in accordance with the state of the fuel injection device, it is possible to prevent an excessive current value from being given. By not giving an excessive current value, the response delay time of the valve closing can be reduced when the injection pulse width is small, and the range of the injection amount in which the relationship between the injection pulse width and the injection amount is a straight line, Can be expanded to the smaller side.

  Further, in order to suppress the individual variation in the injection amount of the fuel injection device 840 or the fuel injection device 2305 of each cylinder in the intermediate lift operation, the individual valve opening start timing Ta ′ and the valve opening completion timing Tb detected by the driving device. Based on this information, a method of controlling the boosted voltage application time Tp or the peak current value Ipeak so that the actual injection period (TB-Ta ′) matches is effective. In this case, the minimum injection amount in the intermediate lift operation is the boost voltage application time Tp, that is, the time during which the switching elements 805 and 806 are energized, to the mover 102 or the mover 1902 by the current supplied to the solenoid 105. It is determined by the stored kinetic energy. Thereafter, a voltage cutoff time T2 for decelerating the mover is provided, and the voltage cutoff time T2 and the holding current value lh are set based on the information of the valve opening completion timing Ta and the valve closing completion timing Tb stored in the drive device. Then, until the valve body 114 or the valve body 1907 reaches the target lift, control is performed so that the valve closing completion timing Tb and the displacement amount of the valve body 114 or the valve body 1907 increase as the injection pulse increases. Further, by adjusting the voltage cutoff time T2 and the holding current value lh based on the detection information, when the valve body 114 or the valve body 1907 reaches the target lift, the speed of the valve body 114 or the valve body 1907 is reduced. Since the bounce of the movable element 102 or the movable element 1902 caused by the collision of the movable element 102 or the movable element 1902 with the fixed core 107 can be reduced, the injection amount after the timing when the target lift is reached from the intermediate lift region is positive. Correlation is established, and the injection amount can be continuously controlled by increasing or decreasing the injection pulse width T.

  As described above, in an engine in which the drive current waveform and the drive pulse width Ti are adjusted and given to each fuel injection device by the ECU, the drive current waveform and the drive pulse according to the manufacturing variation and state of each fuel injection device. Therefore, the ECU 05A reads the valve opening start timing, the valve opening completion timing, and the valve closing completion timing as the state of each fuel injection device.

  When reading the valve opening start timing, the valve opening completion timing, and the valve closing timing of each fuel injector, it is preferable to operate each fuel injector with a drive current waveform that makes it easy to detect the timing of the on-off valve. However, in a drive current waveform that is easy to detect, the linear relationship between the ejection pulse width and the ejection amount may not necessarily be widened.

  For this reason, it is good for ECU05A to have the motive power which sets the drive current waveform for reading the state of a fuel-injection apparatus. For example, the fuel injection device for each connected cylinder using a drive current waveform for reading the behavior of the valve body 114 in a situation where the injection amount is not necessarily minimum, such as during warm-up after the engine is started. The valve opening start timing, the valve opening completion timing, and the valve closing completion timing are detected and recorded in the memory in the ECU 05A. Alternatively, in the split injection condition for dividing the fuel injection in one intake / exhaust stroke, the injection is performed under the condition for causing the valve body 114 or the valve body 1907 to reach the target lift and the condition for performing the intermediate lift operation. It is effective to be able to acquire the detection information of the valve opening start timing and the valve closing completion timing necessary for correcting individual variations in the injection amount of the fuel injection device of each cylinder a plurality of times.

  Based on the recording information of the drive device, the ECU 05A can control and inject a smaller injection amount by adjusting the drive current waveform and the drive pulse width given to each cylinder.

  Thus, by setting a drive waveform for reading the state of the fuel injection device and recording the state of the fuel injection device in a specific engine operation state, the injection amount can be corrected and controlled. The minimum injection amount can be reduced. Further, in the method of performing such learning, it is possible to monitor the state of deterioration of the fuel injection device over time, so that even if the operation of the fuel injection device changes due to deterioration over time, the controllable injection amount The minimum value can be kept small.

  Note that the specific engine operating state is not only during warm-up after engine start, but also during idling, during the engine start process, and several cycles of intake / exhaust stroke after engine key-off, etc. A state in which the rotation speed and load can be adjusted regardless of the operation of the accelerator pedal and the injection amount is not extremely small is a particularly easy period.

  In addition, the valve opening start timing, the valve opening completion timing, and the valve closing timing of the fuel injection device are recorded in the memory in the ECU in this way, and the injection pulse width Ti and the drive current waveform are corrected for each fuel injection device Even in the case of the method performed every time, the timing of the valve operation may be further detected for each injection and reflected in the pulse width command value from the ECU. In particular, when the valve closing completion timing, which is a valve closing operation, is detected by detecting the voltage between the terminals of the solenoid 105 of the fuel injection device or the potential difference between the ground potential (GND) side terminal of the solenoid 105 and the ground potential. Since this can be detected without using a detection-dedicated waveform, it is possible to detect the valve closing completion timing for each fuel injection. By feeding back the detection result to the injection pulse width at the next injection, the control accuracy of the fuel injection amount can be further improved, and the change in the operation of the fuel injection device due to engine temperature, vibration, etc. can be corrected. It becomes like this.

  In this way, as a result of being able to control to a smaller injection amount and use it in an internal combustion engine, it becomes possible to control the injection amount to a smaller injection amount and perform fuel injection, for example, idle stop, etc. This makes it possible to perform combustion at a low load such as recovery from a fuel cut, and it is easy for the engine to have low fuel consumption. In addition, since A / F can be brought close to the target value, gases such as HC and NOX contained in the exhaust can be suppressed. Furthermore, since the fuel injection amount is reduced, the fuel injected during one stroke of the engine can be divided and injected in a plurality of times in a low load region. As a result, the penetrating power of the spray is reduced or the control of forming the air-fuel mixture is facilitated, the fuel adhering to the wall surface of the combustion chamber is suppressed, the homogeneity of the air-fuel mixture is made uniform, and the fuel rich region is Since it can reduce, it can lead to reduction of the discharge amount of soot which is a part of PM (particulate matter) and PN (PM particle number concentration).

  Next, with reference to FIGS. 33 and 34, description will be given of another method for detecting the valve opening start timing which is a factor of the individual variation in the configuration and operation of the fuel injection device and the injection amount in the sixth embodiment. 33, the same symbols are used for parts equivalent to those in FIG.

  First, the configuration and basic operation of the fuel injection device in the sixth embodiment will be described with reference to FIG. FIG. 33 is a diagram showing a configuration of a longitudinal sectional view of the fuel injection device. The fuel injection device shown in FIG. 33 is a normally closed electromagnetic valve (electromagnetic fuel injection device). In a state where the solenoid 105 is not energized, the valve body 3614 is valved by a spring 110 which is a first spring. It is biased toward the seat 118 and is in close contact with the valve seat 118. In this valve closed state, the mover 3602 is biased toward the fixed core 107 (in the valve opening direction) by a zero position spring 3612 as a second spring, and is provided at the end of the valve body 3614 on the fixed core side. It is in close contact with the regulated portion 3614a. In this state, there is a gap between the mover 3602 and the fixed core 107. A lot guide 3613 for guiding the lot portion 31614b of the valve body 3614 is fixed to a nozzle holder 3601 constituting the housing. The valve body 3614 and the mover 3602 are configured to be relatively displaceable, and are contained in the nozzle holder 3601. The lot guide 3613 constitutes a spring seat for the zero position spring 3612. The force by the spring 110 is adjusted at the time of assembly by the pushing amount of the spring retainer 3624 fixed to the inner diameter of the fixed core 107. The urging force of the zero position spring 3612 is set smaller than the urging force of the spring 110.

  In the fuel injection device, the fixed core 107, the movable element 3602, and the housing 3603 constitute a magnetic circuit, and a gap is provided between the movable element 3602 and the fixed core 107. A magnetic aperture 3611 is formed in a portion corresponding to the gap between the mover 3602 and the fixed core 3606 of the nozzle holder 3601. The solenoid 105 is attached to the outer peripheral side of the nozzle holder 101 while being wound around the bobbin 104.

  A lot guide 115 is provided in the vicinity of the end of the valve body 114 opposite to the regulating portion 114 a so as to be fixed to the nozzle holder 101. The rod guide 115 may be configured as the same part as the orifice cup 116. The valve body 114 is guided in movement in the valve axis direction by two rod guides, a first rod guide 113 and a second rod guide 115.

  An orifice cup 116 in which a valve seat 118 and a fuel injection hole 119 are formed is fixed to the tip of the nozzle holder 101, and an internal space (fuel passage) in which the movable element 3602 and the valve body 3614 are provided is externally provided. It is sealed.

  Fuel is supplied from the upper part of the fuel injection device, and the fuel is sealed by a seal portion formed at the end of the valve body 3614 opposite to the regulating portion 3614a and the valve seat 118. When the valve is closed, the valve body is pushed in the closing direction by a force corresponding to the seat inner diameter at the valve seat position by the fuel pressure.

  When a current is passed through the solenoid 105, a magnetic flux is generated between the mover 3602 and the fixed core 107, and a magnetic attractive force is generated. When the magnetic attractive force acting on the mover 3602 exceeds the sum of the load by the spring 110 and the force by the fuel pressure, the mover 3602 moves upward. At this time, the movable element 3602 moves upward together with the valve element 3614 while being engaged with the restricting portion 3614a of the valve element 3614, and moves until the upper end surface of the movable element 3602 collides with the lower surface of the fixed core 107. At this time, if the current supply to the solenoid 105 is stopped before the valve body 3614 reaches the target lift after the valve body 3614 starts to be displaced, an intermediate lift operation is performed.

  As a result, the valve body 3614 is separated from the valve seat 118 and the supplied fuel is injected from the plurality of fuel injection holes 119.

  When the energization to the solenoid 105 is cut off, the magnetic flux generated in the magnetic circuit disappears and the magnetic attractive force disappears. When the magnetic attractive force acting on the mover 3602 disappears, the valve body 3614 is pushed back to the closed position in contact with the valve seat 118 by the load of the spring 110 and the force of the fuel pressure.

  In the state where the valve body 3614 is stationary at the target lift position, that is, in the valve open state, the movable element 3602 and the fixed core 107 collide with one or both of the movable element 3602 and the fixed core 107 on the annular end surface facing each other. The protrusion part of the part is provided. Further, a gap is formed between the opposing end surfaces of the movable element 3602 and the fixed core 107 in the valve open state by the protrusion, except for the protrusion. One or more fuel passages through which the fluid can move are provided. In the operation in which the valve body 3614 is pushed back to the closed position, the mover 3602 moves together while being engaged with the restricting portion 114a of the valve body 114.

  In the fuel injection device of the present embodiment, the valve body 114 and the movable element 3602 are divided into the moment when the movable element 3602 collides with the fixed core 107 when the valve is opened and the moment when the valve element 3614 collides with the valve seat 118 when the valve is closed. By causing relative displacement for a very short time, there is an effect of suppressing the bounce of the movable element 3602 to the fixed core 107 and the bounce of the valve body 114 to the valve seat 118.

  By configuring as described above, the spring 110 biases the valve body 114 in the direction opposite to the direction of the driving force by the magnetic attractive force, and the zero position spring 112 is the biasing force of the spring 110. The mover 3602 is urged in the reverse direction.

  Next, a method for detecting the valve opening start timing when the fuel injection device of FIG. 33 is used will be described with reference to FIG. FIG. 34 shows the voltage Vinj between the terminals of the solenoid 105, the drive current supplied to the solenoid 105, the current value under the condition that the valve element does not open, the difference between the current values of each individual, and the time after the valve displacement and the injection pulse are turned on. It is the figure which showed the relationship. In the drawing of the drive current and the valve displacement, the profiles of the individuals 1, 2 and 3 having different valve opening start timings and the profiles under the condition where the valve element does not start opening are described. From FIG. 33 and FIG. 34, under the condition that the boosted voltage VH is applied and the valve body starts to open at a high current, the magnetic flux on the suction surface is close to saturation. The change in induced electromotive force is small, and as a result, the change in drive current is also small. Further, in the fuel injection device of FIG. 33, since the mover 3602 is stationary, the valve opening is started gradually when the force in the valve opening direction exceeds the force in the valve closing direction. Since the change in acceleration at is small, the change in drive current is small even when the valve opening start timing changes. In such a structure of the fuel injection device, the drive current under the condition that the valve element 3714 does not start opening is stored in the CPU 801 or the IC 802, and the stored drive current and the condition under which the valve element 3714 starts to open the valve. By taking or comparing the difference with the drive current of the fuel injection device of each cylinder at, it is possible to detect a slight change in the drive current accompanying the start of valve opening. At this time, since the change of the current difference accompanying the start of valve opening of the valve body 3714 also rises gently, by setting the threshold value in the current difference, the timing exceeding the threshold value is detected as the valve opening start timing. The valve opening start delay time from when the injection pulse is turned on until the valve opening start timing is preferably stored in the CPU 801 or the IC 802. The drive current (hereinafter referred to as reference current) is acquired under the condition that the valve element 3714 does not start to open under the condition that the fuel pressure supplied to the fuel injection device is high and the differential pressure acting on the valve element 3714 is large. And it is good to detect for every fuel-injection apparatus of each cylinder. The profile of the drive current flowing through the solenoid 105 is affected by individual variations such as the resistance value of the solenoid 105 and the inductance of the magnetic circuit. Therefore, it is possible to accurately detect the valve opening start timing by storing the drive current under the condition that the valve opening does not start for each fuel injection device of each cylinder and taking the difference from the drive current of each fuel injection device. Thus, the correction accuracy of the injection amount can be increased. In addition, when the capacity of the storage memory mounted on the CPU 801 to IC 802 is small, the memory area that can be stored is limited, and therefore, the reference current and the drive current are stored in the detection of the valve opening start timing of a certain cylinder. It is preferable that the reference current and the drive current are erased once in stages and the reference current and the drive current for detecting the valve opening start timing of the fuel injection device for the next cylinder are stored. As a result, the memory usage capacity of the CPU 801 to IC 802 can be reduced, and the sampling rate of the data point sequence to be stored can be made finer, so that the detection accuracy of the valve opening start timing can be increased. Further, according to the method in the sixth embodiment, it is possible to control the valve body 3614 to reach the target lift using a large drive current, which is effective when the fuel injection device is operated under a high fuel pressure condition. is there.

  In the closed state where the valve body 3614 is in contact with the valve seat 118, a differential pressure that is the product of the seat area and the fuel pressure acts on the valve body 3614. Accordingly, when the fuel pressure increases, the differential pressure acting on the valve body 3614 also increases, so that the valve opening start timing of the valve body 3614 is delayed. Since the differential pressure can be calculated as the product of the seat area and the fuel pressure, the relationship between the fuel pressure and the valve opening start timing is a substantially linear relationship. Are stored in the CPU 801 to IC 802, and the relationship between the fuel pressure and the valve opening start timing is made into a function, so that the valve opening start timing for each fuel injection device of each cylinder and the opening when the fuel pressure changes. The valve start timing can be calculated by the ECU 120. The injection period during which the valve element 3614 is displaced can be obtained under the intermediate lift conditions from the information on the valve opening start timing or the valve opening start delay time and the information on the valve closing completion timing, and the injection periods coincide with each other. In this way, by controlling the drive current, the injection amount at the intermediate lift can be controlled, so that a fine injection amount control is possible.

  Next, a fuel injection timing correction method according to the seventh embodiment will be described with reference to FIG. The seventh embodiment is an injection timing control method that can be used in combination with the injection amount control method described in the first to fourth embodiments. The horizontal axis in FIG. 35 shows the timing from the top dead center (TDC) to the bottom dead center (BDC) of the engine piston from the intake stroke to the compression stroke. Further, FIG. 35 shows information on the valve opening start delay time of each individual detected by the ECU with respect to the individual 1, the individual 2, and the individual 3 having different valve opening start timings Ta ′ when performing the divided injection twice. 6 is a graph showing the relationship between the injection pulse and the injection period Tqr during which fuel is injected when the injection timing is controlled. From FIG. 35, it is preferable to inject fuel in the intake stroke during the transition from TDC to BDC from the viewpoint of improving the homogeneity of the air-fuel mixture by improving the flow of the injected fuel and air and reducing the adhesion of the piston. When the injection pulse Ti is input to the drive circuit at the same timing on the basis of TDC in individuals with different valve opening start timings Ta ′, the timing at which fuel injection starts varies from individual to individual, resulting in a homogeneity distribution of the mixture. Fluctuations occur and the injection start timing is delayed, which increases the fuel piston adhesion. There may be an increase in unburned particles including soot. By matching the fuel injection timing for each cylinder, the variation factor from when the fuel is injected until it is mixed with air to form an air-fuel mixture is suppressed, so the homogeneity of the air-fuel mixture for each cylinder Fluctuations can be suppressed, and exhaust performance and fuel consumption can be improved. The individual valve opening start delay time varies with the variation of the valve opening start timing Ta ′ for each of the individual 1, the individual 2, and the individual 3, but for the individual 2 with a long valve opening start delay time, the valve opening start delay time is standard. The injection pulse Ti is output at timing t3901 for the individual 1 and the injection pulse Ti is output at timing t3903 for the individual 2 with a short valve opening start delay time, so that the fuel injection start timing t3904 is output for each individual. Can match. In particular, during split injection in which fuel injection is performed a plurality of times during the intake and exhaust strokes, the time required for driving the valve body 114 to valve body 1907 to reach the target lift position is shorter than in the case of single injection. Therefore, the transient behavior of the valve body 114 to the valve body 1907 at the intermediate lift becomes the dominant factor that determines the fuel injection amount. Also, at the time of split injection, a difference in the injection start timing for each cylinder occurs for the number of split injections, so that an increase in fuel wall surface attachment due to a change in the injection timing and a region where the fuel in the air-fuel mixture is rich may occur. In this case, unburned particles including soot may increase and exhaust performance may deteriorate.

  According to the method in the present embodiment, the homogeneity of the air-fuel mixture for each cylinder can be brought close to the same state by adjusting the timing at which the injection pulse width Ti is supplied for each cylinder. Since combustion particles can be suppressed, exhaust performance can be improved. Further, the injection period Tqr during which the fuel is injected can be matched by correcting the setting of the drive current and the width of the injection pulse Ti for each cylinder using the control methods of the first, third, and fourth embodiments. By the method described above, the injection start timing and the injection end timing t3904 can be matched for each individual (each cylinder), so that variation in the mixture of the cylinders can be suppressed, and the PN and PM contained in the exhaust gas can be reduced. It can be greatly suppressed.

  In the first to seventh embodiments described above, a case where fuel injection in one combustion cycle is performed in a plurality of times will be described with reference to FIGS. 36 and 37. FIG. FIG. 36 shows the relationship between the injection pulse, the injection period, and the valve lift when the fuel injection in one combustion cycle is divided into a plurality of times. FIG. 37 shows the change in engine speed from cranking to the fast idle section.

  As shown in FIG. 36, in the present embodiment, the above-described fuel injection devices 840 and 2305, the CPU 801, and the IC 802 are used to execute divided injection in which fuel is divided and injected multiple times during one combustion cycle. . In FIG. 36, in particular, fuel is injected in four times in the intake stroke from the top dead center (TDC) to the bottom dead center (BDC) of the piston. After passing through the bottom dead center (BDC), the process proceeds to the compression stroke. The four fuel injections include a valve lift 4001v based on the injection pulse 4001p, a valve lift 4002v based on the injection pulse 4002p, a valve lift 4003v based on the injection pulse 4003p, and a valve lift 4004v based on the injection pulse 4004p. .

  In the present embodiment, the amount of fuel injected by the valve lift 4001v is 15% (division ratio: 0.15) of the total amount injected by four fuel injections. The amount of fuel injected by the valve lift 4002v is 70% of the total amount (division ratio: 0.70). The amount of fuel injected by the valve lift 4003v is 10% of the total amount (division ratio: 0.10). The amount of fuel injected by the valve lift 4004v is 5% of the total amount (division ratio: 0.05). When the fuel injected by the valve lifts 4001v to 4004v is combined, the injection pulse width is set so as to be 100% (division ratio: 1) with respect to the target injection amount required during one combustion cycle.

  Of the four fuel injections, the valve lift 4002v is a full lift, and the other valve lifts 4001v, 4003v, and 4004v are half lifts. Here, the full lift is a state in which the valve body is displaced (opened) to a state where the displacement of the valve body is regulated by the regulating unit that regulates the amount of displacement of the valve body, and the valve body reaches the regulating unit. The valve open state at the time. In this state, fuel injection is performed in a state where the valve body has reached the lift position at which the target opening is reached. Moreover, a half lift means the valve opening state by the intermediate opening degree which a valve body does not reach | attain a control part. In this state, fuel injection is performed in a state where the valve element does not reach the lift position at which the target opening is reached. A plurality of divided injections performed in one combustion cycle include a fuel injection by a half lift with a low control accuracy of the injection amount and a fuel injection by a full lift that can control the injection amount with a high accuracy. The total amount of fuel injected by the injection can be made closer to the target injection amount as compared with a case where all the fuel injections are constituted by fuel injection by half lift. Thereby, the drive device of the fuel injection device that can improve the accuracy of the injected fuel injection amount can be provided.

  By the way, as explained in FIGS. 4 and 5, in the fuel injection by the half lift, the change in the fuel injection amount does not become linear with respect to the change in the injection pulse width, but changes nonlinearly. 4 and 5 correspond to conditions such as point 401 and point 501. In particular, as described with reference to FIG. 5, under the half lift condition such as point 501, individual variations of the fuel injection device greatly affect the fuel injection amount. The individual variation of the fuel injection device is noticeable in the variation of the valve opening start timing and the valve closing completion timing of the valve body. The valve opening start timing of the valve element affects the valve closing completion timing, and the earlier the valve opening start timing, the later the valve closing completion timing. That is, the earlier the valve opening start timing, the more the fuel injection amount tends to increase.

  As can be seen from the description in FIGS. 4 and 5, the ratio of the pulse width of the injection pulses 4001p to 4004p to the valve lifts 4001v to 4004v described in FIG. do not do. That is, it is necessary to correct the pulse widths (valve opening start timing and valve closing completion timing) of the injection pulses 4001p to 4004p so that the fuel injection amounts in the valve lifts 4001v to 4004v have the division ratio shown in FIG. is there. This correction can be corrected by using a correction method suitable for the structure of the fuel injection device, as described in the above embodiments.

  In this embodiment, the fuel injection amount of the valve lift 4002v, which is a full lift, is set to 70%. However, the present invention is not limited to this. Also, the number of divisions is not limited to four in the present embodiment. In short, it is necessary to divide the fuel injection in one combustion cycle into a plurality of times, and at least one of the plurality of times must be a fuel injection by a full lift, and the divided fuel injection amount for each time Is the target injection amount required during one combustion cycle. The number of divisions and the divided fuel injection amount may be determined so as to satisfy this condition.

  In FIG. 36, all the divided injections are performed while the piston moves from the top dead center (TDC) to the bottom dead center (BDC) in the intake stroke, but the divided injection is compressed following the intake stroke. It may be done during the journey. The divided injection performed during the compression stroke is desirably injection at an intermediate opening by half lift. In the compression stroke, the piston that has moved to the bottom dead center in the intake stroke moves toward the top dead center. For this reason, the space in which fuel is injected by the piston rising toward the top dead center becomes narrower. In the fuel injection by the full lift, the reach distance (penetration) of the injected fuel spray tends to be long. When fuel injection by full lift is performed in this situation, the piston approaches the injected fuel spray, and the fuel spray easily adheres directly to the piston. This fuel spray is not preferred for combustion in internal combustion engines.

  Therefore, energization of the solenoid 105 for performing fuel injection at the target opening by full lift is performed during the intake stroke in one fuel cycle, and the solenoid 105 for performing fuel injection at intermediate opening by half lift is performed. The energization may be performed at a timing before and after fuel injection at the target opening. In particular, fuel injection at the target opening by full lift is performed when the piston is near the bottom dead center, and further when the piston is near the bottom dead center and descending toward the bottom dead center. Is desirable.

  Further, it is preferable to perform a smaller amount of fuel injection at an early stage of the intake stroke. In addition, when fuel injection is performed also in the compression stroke, it is preferable to perform a smaller amount of fuel injection when the fuel injection is performed at a later time of the compression stroke. As a result, even when the piston is located near the top dead center, the reach of the fuel spray to be injected is short, so that it is possible to avoid the adhesion of the spray to the piston and to ensure a longer period during which fuel can be injected. can do. The reason why the fuel injection at the intermediate opening is shorter than the fuel injection at the target opening is that the injected fuel inevitably becomes smaller, so the flow rate of the air flow formed by the injection is This is because the fuel becomes larger and the relative speed is lowered, and the injected fuel is easily subjected to air resistance. By utilizing this effect, in the fuel injection at the intermediate opening, the penetration can be arbitrarily controlled by changing the energization time (injection pulse width) to the solenoid and the actual injection period that changes depending on the injection interval. .

  It is preferable to perform fuel injection at an intermediate opening by half lift a plurality of times during one combustion cycle, and to change the energization time to the solenoid, that is, the injection pulse width, in the fuel injection at the intermediate opening. Thereby, since data for a plurality of energization times can be acquired during one combustion cycle, it is possible to acquire and learn data in a short time. The energization time to the solenoid may be set by the high voltage application time Tp and the peak current value.

  When performing the split injection described above, either the valve closing completion timing when the valve element contacts the valve seat under the condition of the intermediate opening, or the valve opening start timing when the valve element is separated from the valve seat by applying a voltage to the solenoid One or both of them may be detected by using a change in the current or terminal voltage of the solenoid, and a function of changing the energization time of the solenoid for each fuel injection device of each cylinder may be provided in the drive device of the fuel injection device. By varying the energization time of the solenoid, it is possible to correct variations in either or both of the valve closing completion timing and the valve opening start timing based on individual variations in the fuel injection device. Injecting an accurate amount of fuel enables high-precision fuel injection.

  Alternatively, when performing the above-described divided injection, the valve closing completion timing when the valve body contacts the valve seat under the condition of the intermediate opening, and the valve opening start timing when the voltage is applied to the solenoid and the valve body is separated from the valve seat. Either or both are detected using changes in the solenoid current or terminal voltage, and the function of changing the set value of the current waveform supplied to the solenoid for each fuel injector of each cylinder is It may be provided in the apparatus. By changing the setting value of the current waveform supplied to the solenoid, variation in either or both of the valve closing completion timing and valve opening start timing of the valve body based on individual variations in the fuel injection device can be reduced. It can correct | amend and can inject the exact fuel quantity and can implement | achieve highly accurate fuel injection.

  More specifically, the fuel injection device drive device includes a battery voltage source, a booster circuit that boosts the battery voltage, and a first switch element that controls energization / non-energization from the booster circuit to the solenoid of the fuel injection device, A switch element including a second switch element for controlling energization / non-energization from the battery to the solenoid and a third switch element for controlling energization / non-energization between the ground potential side terminal of the solenoid and the ground potential; and a solenoid A first diode for supplying a current to the booster circuit side, a voltage source side terminal of the solenoid, and a ground potential, provided between the ground potential side terminal of the first switch element and the booster circuit side terminal of the first switch element And a second diode for supplying current from the ground potential side to the voltage source side, and the energization time or energization current of the solenoid is varied based on the valve opening start timing And a means for correcting variations in the fuel injection device, a valve opening start timing detecting section for detecting a valve opening start timing at which the valve element separates from the valve seat, and a drive signal generating section for generating a drive signal for driving the switch element. It is good to provide. The target fuel injection device is driven by a magnetic attraction force from a solenoid, and when contacting the valve body, a movable element that urges the valve body in a valve opening direction, and a contact surface between the valve body and the movable element It is good to provide the space | gap for contacting a valve body after a movable element carries out idle running operation | movement with the magnetic attraction force from a solenoid. The valve opening start timing detection unit is configured such that, at the time of fuel injection at the target opening or the intermediate opening, the drive signal generation unit energizes the first switch element and the third switch element to move the mover in the valve opening direction. After the drive and the first switch and the third switch are de-energized to attenuate the energization current of the solenoid, the mover has a gap during the period of energizing the second switch and the third switch. It is preferable to detect the change in the speed or acceleration of the mover due to the idling operation and contact with the valve body based on the value of the current flowing through the solenoid to detect the valve opening start time.

  At this time, in a period in which the valve body is driven at an intermediate opening and the second switch and the third switch element are energized, the solenoid is accelerated after the mover is accelerated to the valve opening method or the valve body is opened. It is preferable to stop the supply of drive current to. When the valve element is operated at an intermediate opening degree with a fuel injection device having a function for the movable element to idle, the valve element is opened using the kinetic energy of the movable element. At this time, if the target opening for injecting the required injection amount is small, even if the supply of drive current to the solenoid is stopped before the valve body starts to open, the kinetic energy of the mover The valve body can perform a valve opening motion. On the other hand, in a fuel injection device in which the mover does not perform idle running, it is preferable to stop the supply of drive current to the solenoid after the valve element is opened.

  (1) In the condition of performing injection at the target opening in the combustion cycle, the first switch element and the third switch element are energized, the drive current is supplied to the solenoid, and the valve body is opened. The third switch is de-energized or the third switch element and the first switch element are de-energized to attenuate the drive current, and then the second switch element and the third switch element are energized. When the battery voltage is applied to the solenoid and the mover or valve body reaches the target opening during the period in which the battery voltage is applied, the change in the acceleration of the mover due to reaching the target opening By detecting the change, a valve opening completion timing when the valve body reaches the target opening is detected, and a detection unit for detecting the valve opening completion timing for each fuel injection device of each cylinder is provided. Multiply by correction factor Thus, the start of valve opening may be estimated. Thereby, the valve opening completion can be detected under the condition of full lift, and the valve opening start timing can be estimated from the detected information. Even in a fuel injection device having a structure in which it is difficult to detect the valve opening start timing with a sensor, the valve opening start timing information can be obtained.

  In addition, it is preferable that the information acquisition timing and learning timing regarding the valve closing completion timing and the valve opening start timing are as follows. That is, during the operation, the valve closing completion timing and / or the valve opening start timing at the intermediate opening of the fuel injection device of each cylinder is detected and stored, and the next operation after the operation is finished In the period from the engine start to the idling operation where the engine speed reaches a certain rotation, either one or both of the valve closing completion timing and the valve opening start timing of the fuel injection device of each cylinder is detected in advance. The stored valve closing completion time and / or valve opening starting time are compared with the detected valve closing completion time and / or valve opening starting time, and the deviation is driven in advance. When the threshold value given to the device is exceeded, either or both of the detected valve closing completion timing and valve opening start timing are stored in the ECU, and the fuel injection of each cylinder is performed before and after the idling operation. The energization time or current waveform of the apparatus may vary. In this way, by comparing either or both of the valve closing completion time and valve opening start time during the previous operation and the next operation, it is possible to replace the fuel injection device during engine maintenance. Even if it exists, it becomes possible to detect automatically the change by having changed the fuel-injection apparatus by ECU. Accordingly, it is possible to suppress correction of an erroneous injection amount that occurs by using the valve closing completion timing or the valve opening start timing at the previous operation, and the robustness of the injection amount compensation can be improved.

  Alternatively, during the period from the start of the engine to the idle operation in which the engine speed reaches a certain rotation, either or both of the valve closing completion timing and the valve opening start timing of the fuel injection device of each cylinder are detected, and the idle Either one or both of the energization time and current waveform of the fuel injection device of each cylinder may be changed after operation.

  As shown in FIG. 37, when the internal combustion engine is started from a cold state (in the case of cold air start), the engine speed increases from cranking to the fast idle period until the engine speed reaches a constant speed. There is a period to do. In the fast idle section, when the injected fuel adheres to the wall surface, the fuel is hard to vaporize and PM and PN tend to increase. Therefore, in the fast idle section, it is necessary to suppress penetration in order to reduce the adhesion of fuel to the piston and cylinder wall surfaces. In order to suppress the penetration, it is necessary to divide and inject the fuel, so it is necessary to inject a minute amount of fuel with high accuracy. It is desirable to acquire and learn information related to the valve closing completion timing and the valve opening starting timing in the increase period of the rotation speed before reaching the fast idle section. As a result, stable fast idle operation is possible, and fuel consumption is improved.

  As a method of adjusting (correcting) the valve closing completion timing and the actual valve opening period from the valve closing completion to the valve opening start, a booster circuit that boosts the battery voltage is provided so that the valve closing completion timings coincide with each other. The time during which the high voltage is applied from the booster circuit can be adjusted for each fuel injector with respect to the plurality of fuel injectors. In addition, it has a booster circuit that boosts the battery voltage. You may make it adjust for every fuel-injection apparatus with respect to several fuel-injection apparatus. The battery voltage source has a function of controlling the current value flowing through the solenoid within a predetermined range set in the driving device by controlling the energization time of the battery voltage source after applying a high voltage from the booster circuit. The fuel injection at the intermediate opening may be controlled by controlling the energization time of the solenoid. As a method of adjusting the valve opening start timing, the timing for energizing the injection pulse so that the valve opening start timing in each cylinder matches is determined based on the detection information of the valve opening start timing acquired in each cylinder. The method of adjusting every is good.

  In the first to seventh embodiments described above, the timing for detecting the valve opening start timing and the valve closing start timing of the fuel injection device for each cylinder in the case where the fuel injection in one combustion cycle is divided into a plurality of times is shown in FIG. A description will be given with reference to FIG. In FIG. 38, the change in engine speed from cranking to the fast idle section in the ninth embodiment is indicated by a solid line, and the pressure of the fuel pipe upstream of the fuel injection device (hereinafter referred to as rail pressure) is indicated by a broken line. doing.

  The difference between the ninth embodiment and the eighth embodiment is a difference in handling method when the engine operating state is different. In this embodiment, after the transition from the initial combustion due to engine cranking to normal combustion, until the rotation speed converges to a certain value, the idling operation is performed by the increase in the engine rotation speed due to engine blow-up. The correspondence method in the driving | running state with the area 3801 which overshoots the rotation speed in is described. The engine is blown up to increase the temperature of the catalyst attached to the exhaust pipe, and the amount of exhaust gas such as Nox and HC can be reduced. When this section 3801 occurs, the fast idle start condition may be set in the section 3801 or after the section 3801 ends. By starting the fast idle after the engine speed overshoots and starts to decrease, the combustion stability accompanying the increase / decrease in the speed can be suppressed, and stable engine combustion becomes possible.

  From the initial combustion until the fast idle speed is reached, the pipe pressure upstream of the fuel injection device of each cylinder increases toward the target pressure 3802 set arbitrarily by the ECU. At this time, the piping pressure, that is, the fuel pressure in the rail piping upstream of the fuel device, can be detected by the ECU using the output value of the pressure sensor installed in the fuel piping upstream of the fuel injection device. The high-pressure pump is provided with a pressurizing chamber for pressurizing the pressure. The pressurizing piston attached to the camshaft of the engine compresses the fuel to pressurize the fuel and supplies the high-pressure fuel to the rail piping. . In the high-pressure pump, a flow rate from a low-pressure pump (hereinafter referred to as a feed pump) provided in the fuel tank into the pressurizing chamber is controlled by a suction valve. For example, the intake valve is an electromagnetic valve composed of a coil, a magnetic circuit, and a mover, similarly to the fuel injection device, and controls the flow rate into the pressure chamber by energizing and de-energizing the coil. . When the engine is started, the rail pressure increases in synchronization with the operation of the pressurizing piston or the operation of the intake valve while repeating the increase and the constant value. The valve opening start timing, the valve closing end timing, and the valve opening completion timing of the fuel injection device are greatly affected by the fuel pressure because the fluid force applied to the valve body or the mover varies depending on the fuel pressure. The fuel injection amount is determined by the pressure of the fuel supplied to the fuel injection device and the actual valve opening period of the valve element obtained by subtracting the valve opening start timing from the valve closing end timing of the fuel injection device of each cylinder. The detection of the valve opening start timing or the valve closing end timing is preferably synchronized with the operation of the intake valve of the high-pressure pump. By detecting the valve opening start timing or valve closing end timing after a certain period of time since the intake valve command signal was output, individual information of the fuel injection device of each cylinder can be detected under conditions where the pressure fluctuation is small. The correction accuracy of the injection amount can be improved. The rail pressure is determined by the pressure drop due to the pressurization of the high pressure pump and the fuel injection of the fuel injection device of each cylinder. Therefore, the pressure rise due to the current pressure sensor and the engine speed and the energization timing of the intake valve and the pressurization operation of the high pressure pump It is preferable to detect the valve opening start timing or the valve closing end timing of each cylinder under the condition of a certain target rail pressure. Due to this effect, the pressure information at the time of the next fuel injection in one combustion cycle can be accurately obtained even when the period when the output value from the pressure sensor is taken into the ECU is longer than the energization time of the intermediate opening. It is possible to obtain the detection information of the valve opening start timing or the valve closing end timing under an arbitrary pressure condition calculated by the ECU.

  In addition, at the time of the start of fast idle, detection of either or both of the valve closing completion time and the valve opening starting time for correcting individual variations in the injection amount at the intermediate opening of the fuel injection device of each cylinder. Or learning is complete. Therefore, after the start of fast idle, fuel injection in one combustion cycle is not limited to the combination of fuel injection at the target opening and fuel injection at the intermediate opening, and all fuel injections should be performed at the intermediate opening. Can do. By performing all fuel injections in one combustion cycle at an intermediate opening, the fuel injection amount per injection can be reduced. Therefore, since the contact area with the air can be increased and the penetration of the injected fuel can be reduced, fuel adhesion to the piston and cylinder wall surfaces can be suppressed, and exhaust gases such as PM and PN can be reduced. Further, since the injection amount per injection is small, even if the fuel injection interval at the intermediate opening is reduced, the mixing of air and fuel is promoted, and the effect of improving the homogeneity of the air-fuel mixture is obtained.

  Here, it is preferable to reduce the energization pulse width of the first and last fuel injections where the top dead center (TDC) of the piston of the engine approaches. For example, the energization pulse width when all fuel injections in one combustion cycle are performed at an intermediate opening is 0.15:, 0.2 in order from the first fuel injection, assuming that the energization pulse width required in one combustion cycle is 1. : 0.2: 0.2: Set to 0.15. With this setting, fuel can be prevented from adhering to the piston and cylinder wall surfaces at the beginning of the intake stroke and at the end of the compression stroke, and the mixture of fuel and air can be promoted at a timing when the amount of inflow air is large. The degree of homogeneity can be improved.

  In addition, fuel injection is performed at an intermediate opening even after the end of fast idle, and after a certain period of time or a certain amount of fuel injection has elapsed, the valve opening start timing and valve closing end timing of each cylinder at the intermediate opening are redetected. And it is good to update the learning value memorize | stored in ECU. In this way, by updating the learning value, it is possible to correct the injection amount accurately even when the injection amount of the fuel injection device changes due to deterioration over time, and suppress variation in the injection amount of each cylinder. it can.

101 Nozzle holder
102a mover
102b mover
103 Housing
104 bobbins
105 Solenoid
107 Fixed core
110 Spring
111 Magnetic aperture
112 Return spring
115 lot guide
114 Disc
114a Regulatory Department
114b lot part 1 1 4b
117 Fixed core
116 Orifice cup
118 Valve seat
119 Fuel injection hole
120 ECU
121 Drive circuit
124 Spring retainer
201 gap
204 End face
205 Contact surface of the valve body 114 with the mover 102a
206 Sliding surfaces of mover 102a and mover 102b
207 End face on the valve body 114 side of the mover 102b
210 Contact surface
840 Fuel injector
801 Central processing unit (CPU)
802 IC
805, 806, 807, 831 Switching element
809, 810, 811, 832, 835 diode
808, 812, 813 Current and voltage detection resistors
814 Booster circuit
830 coils
815 Ground potential (GND)
620 operational amplifier
841 Terminal on the ground potential (GND) side of the solenoid
R81, R82, R83, R84 resistors
852, 853 Resistor for VL1 voltage detection
C81, C82 capacitors
860 Active low-pass filter for voltage VL1 detection
861 Active low-pass filter for voltage VL2 detection
1501 Analog differentiation circuit
1901 gap
1902 Second mover
1903 First part
1904 Joint
1905 Vertical hole fuel passage
1906 Horizontal hole fuel passage
1907 Second disc
1908 Second Regulatory Department
1909 Initial position spring
1910 First Regulatory Department
2101 Second gap
2201 Third gap
ds sheet diameter
T13 Back pulse application time
Ti injection pulse width (valve opening signal time)
Ta 'valve opening delay time
Ta valve opening completion delay time
Tb Valve closing completion delay time
Tp Boost voltage application time
T2 Drive voltage cutoff time
VH boost voltage
VB battery voltage
Ipeak Peak current value
Ih Holding current value
Tn dead zone
4001p, 4002p, 4003p, 4004p Injection pulse
4001v, 4002v, 4003v, 4004v valve lift

In order to solve the above problems, a fuel injection device drive device according to the present invention comprises:
In a drive device for a fuel injection device that performs split injection that splits and injects fuel multiple times during one combustion cycle,
During the intake stroke during the one combustion cycle,
The valve body or the mover reaches the restricting portion that restricts the displacement in the valve opening direction of the valve body or the mover to the displacement at which the opening of the valve body becomes the target opening. Fuel injection,
An intermediate opening fuel injection for injecting fuel at an intermediate opening that does not reach the target opening by preventing the opening of the valve body from reaching the target opening;
Fuel injection to include
The fuel injection amount in the target opening fuel injection is set larger than the fuel injection amount in the intermediate opening fuel injection,
The intermediate opening fuel injection is performed before and after the target opening fuel injection.

According to the present invention, in a plurality of split injections performed in one combustion cycle, the target opening fuel injection, viewed contains an intermediate degree of opening the fuel injection, the fuel injection quantity of the target opening fuel injection intermediate By setting the fuel injection amount to be larger than the fuel injection amount in the opening fuel injection, the total amount of fuel injected by the plurality of divided injections can be brought close to the target injection amount. Thereby, the drive device of the fuel injection device that can improve the accuracy of the injected fuel injection amount can be provided.

Claims (1)

A mover driven by a solenoid, a valve body that operates in cooperation with the mover, closes by contacting the valve seat, and opens by leaving the valve seat, and a displacement amount of the mover or the valve body A drive device that drives a fuel injection device including a restriction portion that restricts the fuel injection, and includes a switch element that controls energization to the solenoid, and performs fuel injection during one combustion cycle by switching the energization state of the switch element. In the fuel injection device drive device having a function of driving the fuel injection device so as to be divided into a plurality of times,
During a plurality of divided injections performed during one combustion cycle, fuel injection at a target opening at which the valve body or the mover reaches the restricting portion, and an intermediate opening where the valve body does not reach the restricting portion. The fuel injection device is driven so as to include a fuel injection at a degree.

Images