JP2006258056A - Fuel injection device for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily define fluctuation quantity of injection quantity due to influence of injection in different cylinders. <P>SOLUTION: A common rail 13 and a fuel injection valve 3 connected to the common rail 13 are provided. Auxiliary fuel is injected nearly at intake top dead center, and reference fluctuation quantity of injection quantity of main injection fluctuating caused by injection action of the auxiliary fuel is stored. Fluctuation quantity of main injection is defined based on the reference fluctuation quantity and injection command pulse length is controlled to make main injection quantity a target value based on fluctuation quantity of the main injection. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

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

In an internal combustion engine in which the nozzle chamber of the fuel injection valve is connected to a common rail via a high-pressure line and performs two fuel injections, for example, pilot injection and subsequent main injection, when the pilot injection is performed, The pressure wave generated in the nozzle chamber of the fuel injection valve propagates in the high-pressure line and reaches the common rail.Then, the pressure wave is reflected by the common rail and then travels in the high-pressure rail toward the nozzle chamber. When it is performed, a pulsation of intense fuel pressure is generated in the nozzle chamber. However, if the main injection is performed when the fuel pressure in the nozzle chamber is pulsating violently in this way, the valve opening timing of the needle valve fluctuates and the injection pressure fluctuates. It fluctuates and deviates greatly from the normal amount. In view of this, an internal combustion engine in which correction is made so that the injection amount of the main injection becomes a regular injection amount based on fluctuations in the opening timing of the needle valve and fluctuations in the injection pressure is known (for example, Patent Document 1).
JP 2004-84657 A

  This internal combustion engine has a problem of fluctuations in the injection amount due to the pressure wave generated in the fuel injection valve returning to the fuel injection valve again. However, for example, when auxiliary fuel injection is performed near the intake top dead center and main injection is performed near the compression top dead center, each stroke of the engine is one of the two cylinders separated by 360 degrees in crank angle. The auxiliary fuel injection in the cylinder and the main injection in the other cylinder will overlap in time. In this case, when the auxiliary fuel injection in one cylinder is earlier than the main injection in the other cylinder, the pressure wave generated when the auxiliary fuel is injected in the fuel injection valve of the one cylinder is generated by the high pressure line, the common rail, It propagates through the high pressure line into the fuel injection valve of the other cylinder, and as a result, the main injection amount in the other cylinder fluctuates due to the influence of this pressure wave.

Further, in this case, when the auxiliary fuel injection in one cylinder is slower than the main injection in the other cylinder, the pressure wave generated when the main injection is performed in the fuel injection valve of the other cylinder is a high pressure line, Propagation through the common rail and the high-pressure line into the fuel injection valve of one cylinder results in fluctuations in the amount of auxiliary fuel injection in one cylinder due to the influence of this pressure wave. Thus, when considering the variation of the injection amount, the influence of the pressure wave between the fuel injection valves of different cylinders must be considered. In this case, since the propagation path of the pressure wave in the same fuel injection valve and the propagation path of the pressure wave between the fuel injection valves of different cylinders are quite different, the fluctuation of the injection amount due to the pressure wave between the fuel injection valves of different cylinders In obtaining this, it is impossible to apply the variation in the injection amount due to the pressure wave in the same fuel injection valve as it is.
An object of the present invention is to provide a fuel injection device in which the injection amount is controlled in consideration of the influence of pressure waves between fuel injection valves of different cylinders.

  In order to achieve the above object, according to the present invention, a common rail and a plurality of fuel injection valves connected to the common rail are provided and are performed between the latter half of the exhaust stroke before and after the intake top dead center and the first half of the intake stroke. At least two injections of the first injection and the second injection performed between the second half of the compression stroke before and after the compression top dead center and the first half of the expansion stroke are performed in one cycle of the engine in each cylinder. In two cylinders whose strokes are separated by 360 degrees in crank angle, one of the first injection in one cylinder and the second injection in the other cylinder is performed before the other. In an injection control device for an internal combustion engine in which the amount of variation with respect to the target value of the subsequent injection changes depending on the interval time from the injection of the second injection to the subsequent injection, the rail pressure is a predetermined reference rail pressure. The reference fluctuation amount of the injection after changing along the reference fluctuation pattern as the interval time increases is stored, and the injection fluctuation amount when the rail pressure is not the reference rail pressure is calculated from the reference fluctuation amount. The time axis gain for correcting the interval time is stored when obtaining, and the reference fluctuation amount is multiplied when obtaining the fluctuation amount of the subsequent injection when the rail pressure is not the reference rail pressure from the reference fluctuation amount. The rail pressure gain is stored, and the fluctuation amount of the subsequent injection is obtained from the reference fluctuation amount using the time axis gain and the rail pressure gain, and the injection amount of the subsequent injection is controlled to the target value using this fluctuation amount. Is done.

  It is possible to accurately control the injection amount from the fuel injection valve to the target value in consideration of the influence of the pressure wave between the fuel injection valves of different cylinders.

  Referring to FIG. 1, 1 is a compression ignition type multi-cylinder internal combustion engine body, 2 is a combustion chamber of each cylinder, and 3 is a fuel provided for each cylinder in order to inject fuel into each combustion chamber 2, respectively. The injection valve, 4 is an intake manifold, and 5 is an exhaust manifold. The intake manifold 4 is connected to the outlet of the compressor 7 a of the exhaust turbocharger 7 through the intake duct 6, and the inlet of the compressor 7 a is connected to the air cleaner 8. A throttle valve 9 driven by a step motor is disposed in the intake duct 6. On the other hand, the exhaust manifold 5 is connected to the inlet of the exhaust turbine 7 b of the exhaust turbocharger 7.

  The exhaust manifold 5 and the intake manifold 4 are connected to each other via an exhaust gas recirculation (hereinafter referred to as EGR) passage 10, and an electronically controlled EGR control valve 11 is disposed in the EGR passage 10. On the other hand, each fuel injection valve 3 is connected to a common rail 13 via a corresponding fuel supply pipe 12. Fuel is supplied from the fuel tank 15 into the common rail 13 by an electronically controlled fuel pump 14 with variable discharge amount, and the fuel supplied into the common rail 13 is connected to the corresponding fuel injection valve 3 via each fuel supply pipe 12. To be supplied. A fuel pressure sensor 16 for detecting the fuel pressure in the common rail 13 is attached to the common rail 13, and the fuel pump 14 is configured so that the fuel pressure in the common rail 16 becomes the target fuel pressure based on the output signal of the fuel pressure sensor 16. The discharge amount is controlled.

  The electronic control unit 20 comprises a digital computer and is connected to each other by a bidirectional bus 21. A ROM (read only memory) 22, a RAM (random access memory) 23, a CPU (microprocessor) 24, an input port 25 and an output port 26 are connected. It comprises. The output signal of the fuel pressure sensor 16 is input to the input port 25 via the corresponding AD converter 27. On the other hand, a load sensor 18 that generates an output voltage proportional to the depression amount L of the accelerator pedal 17 is connected to the accelerator pedal 17, and the output voltage of the load sensor 18 is input to the input port 25 via the corresponding AD converter 27. Is done. Further, the input port 25 is connected to a crank angle sensor 19 that generates an output pulse every time the crankshaft rotates, for example, 15 °. On the other hand, the output port 26 is connected to the fuel injection valve 3, the step motor for driving the throttle valve 9, the EGR control valve 11, and the fuel pump 14 through the corresponding drive rotation 28.

  FIG. 2 shows an enlarged view of the fuel injection valve 3 shown in FIG. As shown in FIG. 2, the fuel injection valve 3 includes a needle valve 31 that can be seated on a valve seat 30, a sac chamber 32 formed around the tip of the needle valve 31, and the sac chamber 32 into the combustion chamber 2. An extending nozzle hole 33 and a nozzle chamber 34 formed around the needle valve 31 are provided. The nozzle chamber 34 is connected to the common rail 13 through a high-pressure fuel supply passage extending in the main body of the fuel injection valve 3 and the fuel supply pipe 12, a so-called high-pressure line 35, and the high-pressure fuel in the common rail 13 is this high-pressure fuel. It is supplied into the nozzle chamber 34 via the line 35.

  A pressure control chamber 36 is formed on the top surface of the needle valve 31, and a compression spring 37 that presses the needle valve 31 toward the valve seat 30 is disposed in the pressure control chamber 36. The pressure control chamber 36 is connected on the one hand to the middle of the high-pressure line 35 via an inlet side throttle 38, and on the other hand, a fuel overflow port 41 that is controlled to open and close by an overflow control valve 40 via an outlet side throttle 39. It is connected to. High pressure fuel is always supplied to the pressure control chamber 36 through the throttle 38, and therefore the pressure control chamber 36 is always filled with fuel.

  When the fuel overflow port 41 is closed by the overflow control valve 40, as shown in FIG. 2, the needle valve 31 is seated on the valve seat 30, so that the fuel injection is stopped. At this time, the inside of the nozzle chamber 34 and the inside of the pressure control chamber 36 have the same fuel pressure. When the overflow control valve 40 is opened, that is, when the fuel overflow port 41 is opened, the high-pressure fuel in the pressure control chamber 36 flows out from the fuel overflow port 41 through the throttle 39, and thus in the pressure control chamber 36. The pressure of gradually decreases. When the pressure in the pressure control chamber 36 decreases, the needle valve 31 rises and fuel injection is started from the injection hole 33.

  That is, a throttle 39 is provided between the pressure control chamber 36 and the fuel overflow port 41, and after a while after the overflow control valve 40 is opened due to other delay elements, fuel injection is performed. Be started. Next, when the overflow control valve 40 is closed, that is, when the fuel overflow port 41 is closed, the pressure in the pressure control chamber 36 is gradually increased by the fuel supplied into the pressure control chamber 36 through the throttle 38. After a while after the overflow control valve 40 is closed, the fuel injection is stopped.

In the embodiment according to the present invention, as shown in FIG. 3A, the auxiliary fuel R is injected before and after the intake top dead center, and the main fuel M is injected at the end of the compression stroke. In another embodiment according to the present invention, as shown in FIG. 3B, the auxiliary fuel R is injected before and after the intake top dead center, and the pilot fuel P is injected before the main fuel M is injected at the end of the compression stroke. Be injected. In another embodiment according to the present invention, as shown in FIG. 3C, a plurality of pilot injections P ₁ , P before the main injection M from the latter half of the compression stroke before and after the compression top dead center to the first half of the intake stroke. ₂ is performed, and after the main injection M, a plurality of post injections P ₃ and P ₄ are performed.

  As described above, in the present invention, fuel injection is performed a plurality of times from each fuel injection valve 3 during one cycle of the engine. In this case, fuel injection performed later according to the injection timing, that is, fuel injection performed later, That is, it may be affected by the previous injection. That is, as shown in FIG. 3A, when the auxiliary fuel injection R and the main injection M are separated by nearly 360 degrees in crank angle, the auxiliary fuel injection R, that is, the pressure pulsation generated by the previous injection Is attenuated before the main injection M, that is, the subsequent injection, so the previous injection has little effect on the subsequent injection.

On the other hand, in the case shown in FIG. 3B, the auxiliary fuel injection R which is the previous injection hardly affects the pilot injection P and the main injection M which are the subsequent injections. On the other hand, the pilot injection P and the main injection M are close in injection timing, so the pilot injection P that is the preceding injection has a great influence on the main injection M that is the subsequent injection. Further, in the case shown in FIG. 3C, since the injections are all close to each other, each injection P, P ₂ , M, P ₃ has a great influence on the other injections P ₂ , M, P ₃ , P _4. give.

  Next, the influence of the previous injection on the subsequent injection will be described by taking as an example the case where the previous injection is the pilot injection P and the subsequent injection is the main injection M as shown in FIG.

  In the embodiment of the present invention, the target total injection amount QT is previously stored in the ROM 22 in the form of a map as a function of the depression amount of the accelerator pedal 17, that is, the accelerator opening L and the engine speed N, as shown in FIG. Is stored within. Further, the target main injection amount QM is stored in advance in the ROM 22 in the form of a map as a function of the total injection amount QT and the engine speed N as shown in FIG. Further, the target pilot injection amount QP is also stored in advance in the ROM 22 in the form of a map as shown in FIG. 4B as a function of the total injection amount QT and the engine speed N.

  The injection start timing θM of the main injection M is also stored in advance in the ROM 22 as a function of the total injection amount QT and the engine speed N in the form of a map as shown in FIG. The injection timing θP is also stored in advance in the ROM 22 in the form of a map as shown in FIG. 4B as a function of the total injection amount QT and the engine speed N. Further, a time interval from the previous injection to the subsequent injection, that is, an interval time is set in advance. In the embodiment according to the present invention, the interval time TI from when the pilot injection P is started to when the main injection M is started is a function of the total injection amount QT and the engine speed N as shown in FIG. As a map in advance in the ROM 22. In the embodiment according to the present invention, the target rail pressure in the common rail 13 is set in advance. In general, the target rail pressure increases as the total injection amount QT increases.

  Now, when the needle valve 31 is opened in FIG. 2 and fuel injection is started, the pressure in the nozzle chamber 34 rapidly decreases. Thus, when the pressure in the nozzle chamber 34 rapidly decreases, a pressure wave is generated, and this pressure wave propagates in the high-pressure line 35 toward the common rail 13. This pressure wave is then reflected at the open end of the high-pressure line 35 into the common rail 13, this time this pressure wave is in the high-pressure line 35 with the pressure reversed relative to the average pressure, ie in the form of a high-pressure pressure wave. To the nozzle chamber 34, and a high pressure is temporarily generated in the nozzle chamber 34. For example, if pilot injection is performed, a high pressure is temporarily generated in the nozzle chamber 34 by a reflected wave on the common rail 13 for a while after that.

  On the other hand, when the needle valve 31 is closed, the flow of fuel is suddenly blocked, so that the pressure in the nozzle chamber 34 temporarily rises and a pressure wave is generated. This pressure wave also propagates in the high-pressure line 35, is reflected by the common rail 13, and returns to the nozzle chamber 34.

  Further, the pressure wave propagating into the nozzle chamber 34 is also generated by the opening / closing operation of the overflow control valve 40. That is, if the overflow control valve 40 is opened, the pressure of the fuel overflow port 41 is suddenly reduced, so that a pressure wave is generated. If the overflow control valve 40 is closed, the pressure of the fuel overflow port 41 is increased. A pressure wave is generated due to the rapid rise. These pressure waves propagate through the pair of throttles 39 and 38 into the nozzle chamber 34 to increase or decrease the pressure in the nozzle chamber 34. At the same time, the pressure waves are reflected in the nozzle chamber 34 and reflected to the common rail 13 or Propagates toward the fuel overflow port 41.

  Thus, when the pilot injection P is performed, the fuel pressure in the nozzle chamber 34 pulsates due to the pressure wave generated by the opening / closing operation of the needle valve 31 and the opening / closing operation of the overflow control valve 40. Next, the main injection M is performed when the fuel pressure in the nozzle chamber 34 is pulsating as described above. However, if the main injection M is performed when the fuel pressure in the nozzle chamber 34 is pulsating as described above, the injection amount increases when the fuel pressure in the nozzle chamber 34 increases, and the fuel in the nozzle chamber 34 increases. When the pressure decreases, the injection amount decreases, so the injection amount of the main injection M varies.

  Next, referring to FIG. 6 and FIG. 7, what fluctuation pattern the fuel amount of the main injection M actually changes, and a basic method of injection control performed based on this fluctuation pattern The description will be given first, and then the injection control method used in the present invention will be described.

Referring to FIG. 6, the horizontal axis Ti represents the interval time (msec) from when the pilot injection P is started to when the main injection M is started, and the vertical axis dQ is the target value of the injection amount of the main injection M. Represents the amount of fluctuation (mm ³ ). In FIG. 6, □ indicates when the rail pressure is 48 MPa, ○ indicates when the rail pressure is 80 MPa, and Δ indicates when the rail pressure is 128 MPa. FIG. 6 shows a case where the pilot injection amount is 2 (mm ³ ) and the main injection amount is 20 (mm ³ ). FIG. 6A shows the actual fluctuation amount dQ with respect to the target value of the injection amount of the main injection M with respect to three different rail pressures, and after the pilot injection is performed from FIG. It can be seen that the fuel pressure of the engine rises and falls repeatedly, that is, pulsates.

  By the way, in FIG. 6 (A), the fluctuation pattern of the main injection amount represented by each curve is different in period, that is, the period is short, such as a higher rail pressure, but moves up and down in the same manner. I understand that. As described above, the fuel pressure in the nozzle chamber 34 fluctuates due to a pressure wave propagating between the nozzle chamber 34 and the common rail 13 or between the nozzle chamber 34 and the fuel overflow port 41. The distance between the nozzle chamber 34 and the common rail 13 is constant, and the distance between the nozzle chamber 34 and the fuel overflow port 41 is also constant. Therefore, if the propagation velocity of the pressure wave is constant, the pilot chamber P is performed after the pilot injection P is performed. The fuel pressure generated in 34 pulsates with a fixed fluctuation pattern.

  However, the propagation speed of the pressure wave varies depending on the fuel pressure and the fuel temperature. That is, the propagation velocity of the pressure wave is expressed by the square root of (E / γ) · g, where E is the bulk modulus, γ is the fuel density, and g is the acceleration of gravity. That is, the propagation speed of the pressure wave is proportional to the square root of the bulk modulus E and inversely proportional to the square root of the density γ. By the way, the volume modulus E is proportional to the fuel pressure and inversely proportional to the fuel temperature. On the other hand, the density γ of the fuel is also proportional to the fuel pressure and inversely proportional to the fuel temperature. However, the rate of change of the bulk modulus E when the fuel pressure or fuel temperature changes is much larger than the rate of change of the density γ, and therefore the propagation velocity of the pressure wave is strongly influenced by the change of the bulk modulus E. Accordingly, the propagation speed of the pressure wave increases as the fuel pressure increases and decreases as the fuel temperature increases. That is, the propagation speed of the pressure wave increases as the rail pressure increases.

  Therefore, when the rail pressure increases, the fluctuation cycle of the fuel pressure in the nozzle chamber 34 becomes short. At this time, the fluctuation pattern of the fuel pressure in the nozzle chamber 34 is in the horizontal axis direction in FIG. It fluctuates in a contracted form. Therefore, as shown in FIG. 6A, when the rail pressure increases, the fluctuation amount dQ of the main fuel fluctuates in such a manner that the fluctuation pattern contracts in the interval time axis direction.

  In FIG. 6A, assuming that the rail pressure 80 MPa indicated by ◯ is the reference rail pressure, and the fluctuation pattern of the fluctuation amount dQ of the main injection at this reference rail pressure is the reference fluctuation pattern, the rail pressure 48 MPa indicated by □ is indicated. In other words, when the rail pressure is lower than the reference rail pressure, if the entire variation pattern of the variation amount dQ of the main injection is uniformly contracted in the interval time axis direction with the interval time Ti = 0 as a fixed point, the variation pattern vertical variation timing Coincides with the vertical fluctuation timing of the reference fluctuation pattern, and when the rail pressure indicated by Δ is 128 MPa, that is, when the rail pressure is higher than the reference rail pressure, the fluctuation amount dQ of the main injection is set at the interval time Ti = 0 as a fixed point. When the entire fluctuation pattern is uniformly stretched in the interval time axis direction, the fluctuation pattern's vertical fluctuation timing is above the reference fluctuation pattern. To match the change time. In FIG. 6B, the fluctuation pattern when the rail pressure is 48 MPa is contracted and the fluctuation pattern of the rail pressure is 128 MPa so that the vertical fluctuation period of the fluctuation pattern matches the vertical fluctuation period of the reference fluctuation pattern. The case where it was made to show is shown.

  When the variation pattern at each rail pressure is contracted or expanded in this way, the variation period of each variation pattern can be superimposed on the variation period of the reference variation pattern. On the other hand, as shown in FIG. 6B, the fluctuation amount dQ of the main injection at the same interval time Ti increases as the rail pressure increases. Therefore, in order to normalize the fluctuation pattern at each rail pressure to a common reference fluctuation pattern, the fluctuation pattern at each rail pressure is increased in the vertical axis direction of FIG. Or it is necessary to shrink or elongate in the decreasing direction. FIG. 6C shows a case where the fluctuation pattern in each rail pressure is contracted or extended in the increasing or decreasing direction of the fluctuation amount dQ of the main injection in order to normalize the fluctuation pattern to a common reference fluctuation pattern. As shown in FIG. 6C, when the fluctuation pattern is contracted or extended in the interval time axis direction in accordance with the rail pressure and contracted or extended in the direction of increase or decrease in the main injection fluctuation amount in accordance with the rail pressure. In addition, each variation pattern can be normalized to a common reference variation pattern.

  Thus, when each fluctuation pattern can be normalized to a common reference fluctuation pattern, the fluctuation amount dQ of the main injection at each rail pressure can be obtained from the common reference fluctuation pattern. For example, in FIG. 6A, if 80 MPa is the reference rail pressure, and the fluctuation pattern indicated by a circle when the rail pressure is the reference rail pressure is a common reference fluctuation pattern, the square mark when the rail pressure is 48 MPa. The fluctuation period indicated by is longer than the fluctuation period of the common reference fluctuation pattern. Accordingly, when the fluctuation amount dQ of the main injection when the rail pressure is 48 MPa is obtained from a common reference fluctuation pattern, the fluctuation amount dQ is contracted by contracting the time axis of the interval time Ti shown in FIG. It corresponds to the fluctuation amount dQ on the common reference fluctuation pattern according to the corrected interval time Ti. The degree of contraction at this time matches the contraction rate when contracting so that the fluctuation cycle of the fluctuation pattern when the rail pressure is 48 MPa matches the fluctuation period of the common reference fluctuation pattern.

  On the other hand, in FIG. 6A, the fluctuation cycle indicated by Δ when the rail pressure is 128 MPa is shorter than the fluctuation cycle of the common reference fluctuation pattern. Therefore, the fluctuation amount dQ of the main injection when the rail pressure is 128 MPa. Is obtained from a common reference fluctuation pattern, the fluctuation amount dQ is obtained by extending the time axis of the interval time Ti shown in FIG. 6A and on the common reference fluctuation pattern corresponding to the extended corrected interval time Ti. Is equal to the fluctuation amount dQ. The degree of extension at this time coincides with the extension rate when the fluctuation period of the fluctuation pattern when the rail pressure is 128 MPa is extended so as to coincide with the fluctuation period of the common reference fluctuation pattern.

  Further, as already described with reference to FIG. 6B, the fluctuation amount dQ of the main injection at the same interval time Ti increases as the rail pressure increases. Therefore, assuming that the fluctuation pattern when the rail pressure is 80 MPa is the common reference fluctuation pattern, when the fluctuation amount dQ of the main injection is obtained from this common reference fluctuation pattern, the common reference fluctuation pattern is obtained when the rail pressure is 48 MPa. The fluctuation amount dQ obtained from the above is corrected to decrease, and when the rail pressure is 128 MPa, the fluctuation amount dQ obtained from the common reference fluctuation pattern is increased and corrected.

On the other hand, FIG. 7 (A) shows the main injection quantity of 5 (mm ³ ), 10 (mm ³ ), 20 (mm ³ ), 30 (mm ³ ) and 40 with the rail pressure kept constant at 48 MPa. The fluctuation amount dQ of the main injection when (mm ³ ) is shown. Even if the interval time Ti is the same, if the injection amount of the main injection changes, that is, if the injection period changes, the fluctuation amount dQ of the main injection changes. Also in this case, in order to normalize the variation pattern in each injection amount of the main injection to a common reference variation pattern, the variation pattern in each injection amount of the main injection is changed in the vertical axis direction in FIG. That is, it is necessary to contract or extend in the direction of increasing or decreasing the fluctuation amount dQ of the main injection. FIG. 7B shows a case where the fluctuation pattern in each injection quantity of the main injection is superposed on the reference fluctuation pattern by contracting or extending in the increasing or decreasing direction of the main injection quantity dQ.

Thus, for example, 80 MPa is set as the reference rail pressure, the pilot injection amount is 2 (mm ³ ), the main injection amount is 20 (mm ³ ), the reference injection amount is set, the rail pressure is the reference rail pressure, and the pilot injection amount. If the common reference fluctuation pattern is stored in the ROM 22 in advance using the fluctuation pattern of the fluctuation quantity dQ of the main injection when the main injection quantity is the reference injection quantity as a common reference fluctuation pattern, the stored common From the reference fluctuation pattern, it is possible to determine the fluctuation amount of the main injection when the rail pressure and the main injection quantity are variously changed. This is the basic method of injection control performed based on the fluctuation pattern. As will be described below, in the embodiment according to the present invention, the injection control is performed using such a common reference variation pattern.

  Now, when the fuel pressure in the nozzle chamber 34 fluctuates, the main injection amount fluctuates because the injection pressure fluctuates. However, the fuel pressure in the nozzle chamber 34 changes with respect to fluctuations in the main injection amount. Variations in the valve opening timing and the lift amount of the needle valve 31 based on the variation also have a great influence. In this case, it is the fluctuation of the fuel pressure in the nozzle chamber 34 before the needle valve 31 opens that affects the valve opening timing of the needle valve 31, and the fluctuation of the injection pressure that causes the fluctuation of the main injection amount is the needle. This is a change in injection pressure after the valve 31 is opened, that is, a change in fuel pressure in the nozzle chamber 34 after the needle valve 31 is opened.

  Thus, with respect to fluctuations in the main injection amount, fluctuations in the fuel pressure in the nozzle chamber 34 before the needle valve 31 opens, and fluctuations in the fuel pressure in the nozzle chamber 34 after the needle valve 31 opens. In this case, fluctuations in the fuel pressure in the nozzle chamber 34 before and after the needle valve 31 is opened independently affect fluctuations in the main injection amount. Accordingly, the fluctuation amount of the main injection is a fluctuation amount based on the fluctuation of the fuel pressure in the nozzle chamber 34 before the needle valve 31 is opened, and the fluctuation of the fuel pressure in the nozzle chamber 34 after the needle valve 31 is opened. It is the sum of the fluctuation amount based on it.

  First, the influence of the fuel pressure in the nozzle chamber 34 before the needle valve 31 is opened on the main injection amount will be described. Next, the fuel pressure in the nozzle chamber 34 after the needle valve 31 is opened is the main fuel pressure. The influence on the injection amount will be described.

  That is, the overflow control valve 40 is opened based on a command to start main injection, the fuel pressure in the pressure control chamber 36 gradually decreases, and the pressure difference between the nozzle chamber 34 and the pressure control chamber 36 is constant. When the pressure is exceeded, the needle valve 31 is opened. In this case, when the fuel pressure in the nozzle chamber 34 suddenly increases due to pressure pulsation when the fuel pressure in the pressure control chamber 36 gradually decreases, or the fuel pressure in the pressure control chamber 36 suddenly decreases. The pressure difference between the nozzle chamber 34 and the pressure control chamber 36 becomes a certain value or more, and thus the opening timing of the needle valve 31 is advanced. On the other hand, when the fuel pressure in the pressure control chamber 36 gradually decreases, the fuel pressure in the nozzle chamber 34 suddenly decreases due to pressure pulsation, or the combustion pressure in the pressure control chamber 36 rapidly increases. Then, since it takes time for the pressure difference between the nozzle chamber 34 and the pressure control chamber 36 to become a certain value or more, the valve opening timing of the needle valve 31 is delayed.

  Thus, the valve opening timing of the needle valve 31 is advanced or delayed by the fluctuation of the fuel pressure in the nozzle chamber 34 before the needle valve 31 is opened or the fluctuation of the fuel pressure in the pressure control chamber 36. Or In this case, the main injection amount increases when the valve opening timing of the needle valve 31 is advanced, and the main injection amount decreases when the valve opening timing of the needle valve 31 is delayed. Therefore, when the valve opening timing of the needle valve 31 fluctuates due to the influence of pressure pulsation, the main injection amount fluctuates accordingly.

FIG. 8 shows the relationship between the interval time Ti (msec) from when the pilot injection P is started to when the main injection M is started and the variation Δτ (μmsec) in the valve opening timing of the needle valve 31. . FIG. 8 shows the case where the pilot injection amount is 2 (mm ³ ), □ indicates when the rail pressure is 48 MPa, ○ indicates when the rail pressure is 80 MPa, The mark indicates when the rail pressure is 128 MPa.

  FIG. 8A shows the actual value of the variation Δτ of the valve opening timing of the needle valve 31 at each rail pressure. In FIG. 8B, the rail pressure of 80 MPa is used as the reference rail pressure, and the fluctuation pattern of the valve opening timing of the needle valve 31 at this time is used as the reference fluctuation pattern, and the rail pressure shown in FIG. 8A is 48 MPa and 128 MPa. In this example, the fluctuation pattern is contracted or extended in the interval time axis direction so that the vertical fluctuation period of these fluctuation patterns matches the vertical fluctuation period of the reference fluctuation pattern.

  On the other hand, FIG. 8C shows the fluctuation pattern when the rail pressure shown in FIG. 8B is 48 MPa and 128 MPa in the vertical direction, that is, the valve opening timing of the needle valve 31 so that these fluctuation patterns overlap the reference fluctuation pattern. A case is shown in which the amount of fluctuation Δτ contracts or extends in the increasing or decreasing direction. Thus, it can be seen that the fluctuation pattern of the fluctuation amount Δτ of the valve opening timing of the needle valve 31 can be normalized as shown in FIG. The reference variation pattern shown in FIG. 8C is stored in advance in the ROM 22, and the variation in the valve opening timing of the needle valve 31 in the corresponding fuel injection valve 3 according to the rail pressure based on this reference variation pattern. The quantity Δτ is calculated.

  Specifically, in order to obtain the variation Δτ of the valve opening timing of the needle valve 31 corresponding to the rail pressure shown in FIG. 8A based on the reference fluctuation pattern shown in FIG. First, the interval time Ti is multiplied by the time axis gain IK1 shown in FIG. 9A to obtain the corrected interval time Ti · IK1, and the corrected interval time Ti · IK1 is used to obtain the corrected interval time Ti · IK1 as shown in FIG. The variation amount Δτ of the valve opening time at the reference rail pressure is obtained from the reference variation pattern, and then the variation amount Δτ is multiplied by the rail pressure gain IK2 shown in FIG. 9B corresponding to the rail pressure. A variation amount Δτ of the valve opening timing (= variation amount Δτ · IK2 at the reference rail pressure) is obtained.

  Here, the time axis gain IK1 shown in FIG. 9A will be described. When the rail pressure is lower than the reference rail pressure 80 MPa, for example, 48 MPa shown by □ in FIG. To match the fluctuation period of the reference fluctuation pattern, it is necessary to contract the interval time axis. Therefore, the fluctuation amount Δτ in the interval time Ti when the rail pressure is 48 MPa is the interval time in the reference fluctuation pattern on the reference fluctuation pattern. Appears earlier than Ti. That is, the correction interval time Ti · IK1 to be used for obtaining the fluctuation amount Δτ when the rail pressure is 48 MPa from the reference fluctuation pattern is an interval time shorter than the interval time Ti when the rail pressure is 48 MPa. When the rail pressure is low, as shown in FIG. 9A, the time axis gain IK1 is made smaller than 1.0.

  On the other hand, when the rail pressure is higher than the reference rail pressure 80 MPa, for example, at 128 MPa indicated by Δ in FIG. 8A, an interval is used to make the fluctuation period of the fluctuation pattern coincide with the fluctuation period of the reference fluctuation pattern. It is necessary to extend the time axis. Therefore, the fluctuation amount Δτ in the interval time Ti when the rail pressure is 128 MPa appears on the reference fluctuation pattern at a time later than the interval time Ti in the reference fluctuation pattern. That is, the correction interval time Ti · IK1 to be used for obtaining the fluctuation amount Δτ when the rail pressure is 128 MPa from the reference fluctuation pattern is set to an interval time longer than the interval time Ti when the rail pressure is 128 MPa. When the pressure is high, the time axis gain IK1 is set larger than 1.0 as shown in FIG. That is, the time axis gain IK1 increases as the rail pressure increases as shown in FIG. The time axis gain IK1 is common to the injection effects from all the fuel injection valves 3.

  Since the variation amount Δτ of the valve opening time tends to increase as the rail pressure decreases, the rail pressure gain IK2 increases as the rail pressure decreases as shown in FIG. 9B.

  When the final valve opening timing fluctuation amount Δτ is determined, the valve opening timing command value is corrected based on the fluctuation amount Δτ so that the actual valve opening timing becomes a target value, for example. For example, when the variation amount Δτ of the valve opening timing is positive, the command value of the main injection is corrected so that the main injection start timing is delayed by the variation amount Δτ, and when the variation amount Δτ of the valve opening timing is negative, The command value of the valve opening timing is corrected so that the start timing is advanced by the fluctuation amount Δτ.

  On the other hand, the completion timing of the main injection can be controlled based on the obtained variation amount Δτ of the valve opening timing. This will be described with reference to FIG. FIG. 10 shows the relationship between the injection command pulse for main injection and the lift amount of the needle valve 31. Hereinafter, the injection control will be described by taking as an example the case where the needle valve 31 uses a type of fuel injection valve that does not open to the maximum lift MAX. However, the type of fuel injection valve that opens the needle valve 31 to the maximum lift MAX. The injection control can be performed in the same way even when using.

  When an injection command pulse is issued as shown by A in FIG. 10, the lift amount of the needle valve 31 begins to rise for a while after the injection command pulse is issued as shown by a solid line B in FIG. When the generation of the command pulse is stopped, it begins to descend after a while. On the other hand, if the opening timing of the needle valve 31 is advanced by Δτ due to pressure pulsation in the nozzle chamber 34 before the needle valve 31 is opened, the lift amount of the needle valve 31 at this time is indicated by a broken line C in FIG. At the same speed as the case indicated by the solid line B, and starts decreasing at substantially the same speed as the case indicated by the solid line B. Thus, when the opening timing of the needle valve 31 is advanced, the injection period becomes longer and the maximum lift amount of the needle valve 31 is also increased, so that the injection amount is increased.

  On the other hand, the injection command pulse indicated by D in FIG. 10 shows the case where the generation time of the injection command pulse is shortened by the variation amount Δτ of the valve opening timing, and the broken line E in FIG. Indicates the amount. In FIG. 10, the change in the lift amount indicated by the broken line E and the change in the lift amount indicated by the solid line B are exactly the same. Therefore, when the opening timing of the needle valve 31 is advanced by Δτ, the generation time of the injection command pulse is reduced. If it is shortened by Δτ, the injection amount becomes the target injection amount.

  In the embodiment according to the present invention, the injection command pulse is shortened when the valve opening timing of the needle valve 31 is advanced, and the injection command pulse is extended when the valve opening timing of the needle valve 31 is delayed, thereby targeting the injection amount. The injection amount is controlled.

  Next, the fluctuation amount dQm of the main injection when the valve opening timing of the needle valve 31 fluctuates will be described. FIG. 11 shows the opening timing of the needle valve 31 shown in FIG. 10 from another viewpoint. The change in the lift amount of the needle valve 31 indicated by the injection command pulse indicated by A in FIG. 11 and the solid line B and the broken line C is indicated by the injection command pulse indicated by A, the solid line B and the broken line C in FIG. The change in the lift amount of the needle valve 31 is exactly the same.

  Now, the injection command pulse indicated by D in FIG. 11 shows the case where the generation time of the injection command pulse is extended by the variation amount Δτ of the valve opening timing, and the broken line E in FIG. 11 shows the lift of the needle valve 31 at this time. Indicates the amount. In FIG. 11, the change in the lift amount indicated by the broken line E and the change in the lift amount indicated by the broken line C are exactly the same. Therefore, when the valve opening timing of the needle valve 31 is advanced by Δτ and the generation time of the injection command pulse. The amount of change in the injection amount is the same as when it is extended by Δτ.

  Thus, the variation amount of the injection amount when the valve opening period of the needle valve 31 is advanced by Δτ is the same as the variation amount of the injection amount when the injection period is extended by Δτ. Therefore, the variation amount of the injection amount when the injection period becomes longer by Δτ will be described next.

  FIG. 12 shows the relationship between the injection command pulse length and the injection amount. This relationship is obtained in advance by experiments, and this relationship obtained by experiments is stored in the ROM 22 in advance. As shown in FIG. 12, when the injection command pulse length is short, the opening time of the overflow control valve 40 is too short, so the needle valve 31 does not open, and therefore the injection amount becomes zero. On the other hand, as the injection command pulse length becomes longer, the lift amount of the needle valve 31 increases, and as a result, the injection amount increases exponentially as the injection command pulse length increases.

  FIG. 13 shows the relationship between the slope ΔQ / Δt of the curve shown in FIG. 12 and the injection amount. That is, ΔQ / Δt represents the change amount ΔQ of the injection amount when the injection command pulse length changes by a certain length Δt at a certain injection amount in FIG. As shown in FIG. 13, the increase rate ΔQ / Δt of the injection amount with respect to the increase in the injection command pulse length increases rapidly first and then slowly when the injection amount increases.

  Using the relationship between ΔQ / Δt and the injection amount shown in FIG. 13, the fluctuation amount dQm of the main injection amount when the valve opening timing of the needle valve 31 fluctuates can be obtained. For example, in FIG. 11, when the opening timing of the needle valve 31 is advanced by Δτ, the injection amount at the injection command pulse length indicated by A in FIG. 11 is obtained from FIG. 12, and ΔQ / Δt at this injection amount. Is obtained from FIG. The fluctuation amount of the injection amount when the valve opening timing fluctuates by Δτ is the same as the fluctuation amount of the injection amount when the injection command pulse length changes by Δτ. Therefore, the fluctuation amount dQm of the main injection amount at this time is ΔQ / It can be calculated using Δt. That is, the fluctuation amount dQm of the main injection amount when the valve opening timing fluctuates by Δτ is a value obtained by multiplying ΔQ / Δt by Δτ ((ΔQ / Δt) · Δτ).

  So far, the fluctuation amount of the main injection based on the fluctuation of the fuel pressure in the nozzle chamber 34 before the needle valve 31 is opened has been described. Next, the fuel in the nozzle chamber 34 after the needle valve 31 is opened. The fluctuation amount of the main injection based on the pressure fluctuation will be described.

  The fluctuation amount of the valve opening timing of the needle valve 31 and the fluctuation amount of the entire main injection amount can be detected, but the fluctuation amount of the main injection based on the fluctuation of the fuel pressure in the nozzle chamber 34 after the needle valve 31 is opened. Can not only detect. However, as described above, the fluctuation amount dQm of the main injection based on the fluctuation of the valve opening timing of the needle valve 31 can be calculated. Accordingly, in the embodiment according to the present invention, the inside of the nozzle chamber 34 after the needle valve 31 is opened by subtracting the fluctuation amount dQm of the main injection based on the fluctuation of the valve opening timing of the needle valve 31 from the fluctuation amount of the entire main injection amount. The fluctuation amount dQt of the main injection based on the fluctuation of the fuel pressure is obtained.

FIGS. 14A and 14B show views similar to FIGS. 7A and 7B with respect to the fluctuation amount dQt of the main injection, respectively. 14A shows the case where the pilot injection amount is 2 (mm ³ ) and the rail pressure is standardized to the reference rail pressure of 80 MPa, and the + mark indicates the case where the main injection amount is 5 (mm ³ ). ◇ indicates when the main injection amount is 10 (mm ³ ), Δ indicates when the main injection amount is 20 (mm ³ ), and ○ indicates the main injection amount is 30 (Mm ³ ) is shown, and □ marks show when the main injection amount is 40 (mm ³ ). On the other hand, in 16 (B), the fluctuation pattern when the rail pressure is the reference rail pressure 80 MPa and the main injection amount is 20 (mm ³ ) is used as the reference fluctuation pattern, and the main injection amounts are 5 (mm ³ ) and 10 (mm). ³ ) The fluctuation pattern at 30 (mm ³ ) and 40 (mm ³ ) is shrunk or extended in the vertical direction, that is, in the direction of increase or decrease in the main injection fluctuation amount dQt so that these fluctuation patterns overlap the reference fluctuation pattern. Shows the case.

In order to normalize the fluctuation amount dQt of the main injection to the state shown in FIG. 14A, first, the fluctuation amount of the main injection that differs depending on the rail pressure as in the fluctuation amount shown in FIG. The interval time Ti at each rail pressure is shrunk or extended so that the fluctuation period of dQt overlaps the reference rail pressure, for example, the fluctuation period of 80 MPa. Next, similarly to the variation shown in FIG. 6B, the variation dQt of the main injection that differs depending on the rail pressure is standardized to a variation pattern at a reference rail pressure, for example, 80 MPa. FIG. 14A shows the main injection fluctuation amount dQt for various main injection amounts when the pilot injection amount thus standardized is 2 (mm ³ ) and the rail pressure is the reference rail pressure 80 MPa. . Next, the fluctuation amount dQt of the main injection is standardized with the fluctuation pattern when the main injection quantity is 20 (mm ³ ) as the reference fluctuation pattern as shown in FIG. 14B. The reference fluctuation pattern is stored in the ROM 22 in advance, and the main injection fluctuation amount dQt is calculated based on the stored reference fluctuation pattern.

That is, in order to determine the fluctuation amount dQt of the main injection, first, the interval time Ti is multiplied by the time axis gain FK1 shown in FIG. Using FK1, the fluctuation amount dQt of the main injection at the reference rail pressure 80 MPa and the reference main injection amount 20 (mm ³ ) is obtained from the reference fluctuation pattern shown in FIG. 14B, and then the fluctuation amount dQt of the main injection is obtained. By multiplying the injection amount gain FK3 shown in FIG. 15 (C) corresponding to the main injection amount, the fluctuation amount dQt of the main injection corresponding to the main injection amount shown in FIG. 14 (A) is obtained, and then this main injection Is multiplied by the rail pressure gain FK2 shown in FIG. 15B corresponding to the rail pressure to obtain the final main injection variation dQt (= reference variation pattern). The main injection of variation dQt · FK1 · FK2) obtained from is determined.

  Here, the time axis gain FK1 shown in FIG. 15A is exactly the same as the time axis gain IK1 shown in FIG. 9A. Further, the rail pressure gain FK2 increases as the rail pressure increases as shown in FIG. On the other hand, the injection amount gain FK3 increases as the main injection amount increases as shown in FIG. In this way, the fluctuation amount dQt of the main injection is calculated using these gains FK1, FK2, and FK3.

  Next, the propagation of pressure waves between the fuel injection valves 3 of different cylinders will be described. FIG. 16 shows the fuel injection valve 3 of each cylinder shown in FIG. 1, that is, the first cylinder # 1, the second cylinder # 2, the third cylinder # 3, the fourth cylinder # 4, the fuel supply pipe 12, The common rail 13 is shown. The injection order from each fuel injection valve 3 in the embodiment shown in FIG. 16 is 1-3-3-2.

  As described above, as shown in FIG. 3A, when the auxiliary fuel injection R and the main injection M are separated from each other by approximately 360 degrees in crank angle, the auxiliary fuel injection R has little influence on the main injection M. Absent. However, the pressure wave generated by the auxiliary fuel injection R reaches not only the pressure wave reflected on the common rail 13 after reaching the common rail 13 through the high pressure line 35, but also the other through the other high pressure line 35 after propagating in the common rail 13. There is also a pressure wave that reaches the fuel injection valve 3 of this cylinder. Accordingly, if the main fuel injection M is performed in different cylinders at the same time when the auxiliary fuel injection R is performed in a certain cylinder, the auxiliary fuel injection R affects the main fuel injection M in the different cylinders. .

  In the example shown in FIG. 3A, the auxiliary fuel injection R and the main injection M are separated by approximately 360 degrees in crank angle. In this case, the cylinder in which the auxiliary fuel injection R affects the main injection M is the auxiliary fuel injection. This is a cylinder in which each stroke of the engine is 360 degrees away from the cylinder in which R is performed. That is, in this case, the combination of the cylinders in which the fluctuation of the injection amount due to the influence of the pressure wave becomes a problem is two cylinders in which each stroke of the engine is separated by 360 degrees in crank angle. In a four-cylinder engine as shown in FIG. 16, this cylinder combination is a combination of the first cylinder # 1 and the fourth cylinder # 4, and a combination of the second cylinder # 2 and the third cylinder # 3.

  FIGS. 17A and 17B show that the auxiliary fuel injection R is more time-consuming than the main injection M between the first cylinder # 1 and the fourth cylinder # 4 and between the second cylinder # 2 and the third cylinder # 3. , That is, the case where the preceding injection is the auxiliary fuel injection R and the subsequent injection is the main injection M. The subsequent injection is the interval time Ti after the previous injection is started. Started when it has passed.

  In the example shown in FIG. 17A, the pressure wave generated by the auxiliary fuel injection R in the fourth cylinder # 4 reaches the fuel injection valve 3 in the first cylinder # 1 through the path indicated by the broken line X in FIG. Variations in the injection amount of the main injection M in the first cylinder # 1 are caused. Further, the pressure wave generated by the auxiliary fuel injection R in the first cylinder # 1 reaches the fuel injection valve 3 in the fourth cylinder # 4 through the path indicated by the broken line X in FIG. 16, and the main injection in the fourth cylinder # 4. A variation in the injection amount of M is caused.

  On the other hand, in the example shown in FIG. 17B, the pressure wave generated by the auxiliary fuel injection R in the third cylinder # 3 passes through the path indicated by the broken line Y in FIG. 16 to the fuel injection valve 3 in the second cylinder # 2. And the fluctuation of the injection amount of the main injection M of the second cylinder # 2 is caused. Further, the pressure wave generated by the auxiliary fuel injection R in the second cylinder # 2 reaches the fuel injection valve 3 of the third cylinder # 3 through the path indicated by the broken line Y in FIG. 16, and the main injection of the third cylinder # 3. A variation in the injection amount of M is caused.

  On the other hand, FIGS. 18A and 18B show that the main injection M is greater than the auxiliary fuel injection R between the first cylinder # 1 and the fourth cylinder # 4 and between the second cylinder # 2 and the third cylinder # 3. The case where the injection is performed slightly earlier, that is, the case where the previous injection is the main injection M and the subsequent injection is the auxiliary fuel injection R, and the subsequent injection is the interval time after the previous injection is started. Started when Ti has elapsed.

  In the example shown in FIG. 18A, the pressure wave generated by the main injection M in the first cylinder # 1 reaches the fuel injection valve 3 in the fourth cylinder # 4 through the path indicated by the broken line X in FIG. Variations in the injection amount of the auxiliary fuel injection R of the numbering cylinder # 4 are caused. Further, the pressure wave generated by the main injection M in the fourth cylinder # 4 reaches the fuel injection valve 3 in the first cylinder # 1 through the path indicated by the broken line X in FIG. 16, and the auxiliary fuel injection in the first cylinder # 1. The injection quantity of R is changed.

  On the other hand, in the example shown in FIG. 18B, the pressure wave generated by the main injection M in the second cylinder # 2 reaches the fuel injection valve 3 in the third cylinder # 3 through the path indicated by the broken line Y in FIG. Variations in the injection amount of the auxiliary fuel injection R of the third cylinder # 3 are caused. Further, the pressure wave generated by the main injection M in the third cylinder # 3 reaches the fuel injection valve 3 in the second cylinder # 2 through the path indicated by the broken line Y in FIG. 16, and the auxiliary fuel injection in the second cylinder # 2 The injection quantity of R is changed.

  The injection timing of the auxiliary fuel injection R that affects the main injection M between the two cylinders whose crank angles are separated by 360 degrees in the engine stroke is from the latter half of the exhaust stroke before and after the intake top dead center to the first half of the intake stroke. The injection timing of the main injection M that affects the auxiliary fuel injection R is from the latter half of the compression stroke before and after the compression top dead center to the first half of the expansion stroke. Therefore, when the relationship between the auxiliary fuel injection R and the main injection M shown in FIGS. 17A and 17B and FIGS. 18A and 18B is generally expressed, the second half of the exhaust stroke before and after the intake top dead center. To the first half of the intake stroke and the second injection performed between the second half of the compression stroke before and after the compression top dead center and the second half of the expansion stroke. In two cylinders that are performed during one cycle and each stroke of the engine is separated by 360 degrees in crank angle, one of the first injection in one cylinder and the second injection in the other cylinder is more than the other. Is performed first, and at this time, the previous injection affects the subsequent injection.

  FIGS. 19A and 19B show the case where the preceding injection is the auxiliary fuel injection R in the fourth cylinder # 4 and the subsequent injection is the main injection M in the first cylinder # 1 in FIG. 17A. The relationship between the interval time Ti (msec) and the fluctuation amount Δτ2 (μmsec) of the valve opening timing of the needle valve 31 is shown. In FIG. 17A, the interval time Ti and the needle valve 31 when the previous injection is the auxiliary fuel injection R in the first cylinder # 1 and the subsequent injection is the main injection M in the fourth cylinder # 4. The relationship with the variation amount Δτ2 of the valve opening timing is the same. In the embodiment according to the present invention, the interval time TI from the start of the auxiliary fuel injection R to the start of the main injection M is the total injection amount QT and the engine speed N as shown in FIG. The function is stored in advance in the ROM 22 in the form of a map.

FIGS. 19A and 19B show a case where the auxiliary fuel injection amount is 2 (mm ³ ) and the main injection amount is 15 (mm ³ ). In FIGS. 19A and 19B, ◯ indicates when the rail pressure is 80 MPa, △ indicates when the rail pressure is 128 MPa, and ◇ indicates when the rail pressure is 180 MPa. Show. When the propagation path of the pressure wave becomes a long path passing through the common rail 13 as in FIGS. 19A and 19B, the pressure wave becomes smooth while propagating, and therefore FIGS. 19A and 8A. As can be seen from the comparison, the variation in the valve opening timing of the needle valve 31 is gentler in the case shown in FIG. 19A than in the case shown in FIG.

  FIG. 19A shows the actual value of the variation Δτ2 of the valve opening timing of the needle valve 31 at each rail pressure. In FIG. 19B, the rail pressure of 128 MPa is used as the reference rail pressure, and the fluctuation pattern of the valve opening timing of the needle valve 31 at this time is used as the reference fluctuation pattern, and the rail pressure shown in FIG. 19A is 80 MPa and 180 MPa. The fluctuation pattern is contracted or expanded in the interval time axis direction so that the vertical fluctuation period of these fluctuation patterns matches the vertical fluctuation period of the reference fluctuation pattern, and further, the fluctuation pattern is contracted or extended in the interval time axis direction. A case is shown in which the valve contracts or extends in the vertical direction, that is, in the direction of increase or decrease in the amount of variation Δτ2 of the valve opening timing of the needle valve 31 so as to overlap the reference variation pattern. Thus, even when the pressure wave propagates between the fuel injection valves 3 of different cylinders, the variation pattern of the variation amount Δτ2 of the valve opening timing of the needle valve 31 can be normalized as shown in FIG. 19B. .

  On the other hand, in FIG. 17B, when the previous injection is the auxiliary fuel injection R in the third cylinder # 3 and the subsequent injection is the main injection M in the second cylinder # 2, as shown in FIG. Compared with the case shown in FIG. 17 (A), the propagation path of the pressure wave becomes shorter. Therefore, the variation amount Δτ2 of the valve opening timing of the needle valve 31 in this case is different from the variation pattern shown in FIG. 19 (A). It becomes a pattern. The reference fluctuation pattern as shown in FIG. 19B for the combination of the second cylinder # 2 and the third cylinder # 3 is not particularly shown, but this reference fluctuation pattern is also preliminarily in the form of a function of the interval time Ti. Required by experiment.

  This reference fluctuation pattern is also stored in the ROM 22 in advance together with the reference fluctuation pattern shown in FIG. 19B, and the opening of the needle valve 31 in the corresponding fuel injection valve 3 corresponding to the rail pressure based on these reference fluctuation patterns. A variation amount Δτ2 of the time is calculated.

  Specifically, in order to obtain the variation amount Δτ2 of the valve opening timing of the needle valve 31 corresponding to the rail pressure shown in FIG. 19A based on the reference fluctuation pattern shown in FIG. 19B. First, the interval time Ti is multiplied by the time axis gain IG1 shown in FIG. 20 (A) to obtain the corrected interval time Ti · IG1, and the corrected interval time Ti · IG1 is used to show the corrected interval time Ti · IG1 as shown in FIG. 19 (B). The variation amount Δτ2 of the valve opening time at the reference rail pressure is obtained from the reference variation pattern, and then the variation amount Δτ2 is multiplied by the rail pressure gain IG2 shown in FIG. 20B corresponding to the rail pressure. A variation amount Δτ2 of the valve opening timing (= a variation amount Δτ2 · IG2 at the reference rail pressure) is obtained.

  The time axis gain IG1 is also less than 1.0 when the rail pressure is low, and the time axis gain IG1 is 1 when the rail pressure is high, similarly to IK1 shown in FIG. Greater than .0. That is, the time axis gain IG1 increases as the rail pressure increases as shown in FIG.

  On the other hand, the amount of variation Δτ2 in the valve opening time tends to increase as the rail pressure decreases, so the rail pressure gain IG2 increases as the rail pressure decreases. Since the variation amount τ2 of the valve opening timing is different for each combination of two cylinders in which each stroke of the engine is separated by 360 degrees in the crank angle, the rail pressure gain IG2 as shown by a and b in FIG. Are provided for the combination of the first cylinder # 1 and the fourth cylinder # 4 and the combination of the second cylinder # 2 and the third cylinder # 3, respectively.

  As shown in FIGS. 18A and 18B, the first cylinder # 1 and the fourth cylinder # 4 are also used when the previous injection is the main injection M and the subsequent injection is the auxiliary fuel injection R. The reference variation pattern and the rail pressure gain IG2 are stored in advance in the ROM 22 for the combination of the second cylinder # 2 and the third cylinder # 3.

  Next, the fluctuation amount dQt2 of the main injection based on the fluctuation of the fuel pressure in the nozzle chamber 34 after the needle valve 31 is opened will be described. The fluctuation amount dQt2 of the main injection is also obtained by the above-described method, that is, by subtracting the fluctuation amount dQm of the main injection based on the fluctuation of the valve opening timing of the needle valve 31 from the fluctuation amount of the entire main injection amount.

  FIGS. 21A and 21B show the case where the preceding injection is the auxiliary fuel injection R in the fourth cylinder # 4 and the subsequent injection is the main injection M in the first cylinder # 1 in FIG. The relationship between the interval time Ti (msec) and the fluctuation amount dQt2 of the main injection is shown. In FIG. 17A, the interval time Ti and the fluctuation of the main injection when the previous injection is the auxiliary fuel injection R in the first cylinder # 1 and the subsequent injection is the main injection M in the fourth cylinder # 4. The relationship with the quantity dQt2 is the same.

FIGS. 21A and 21B show a case where the auxiliary fuel injection amount is 2 (mm ³ ) and the main injection amount is 15 (mm ³ ). In FIGS. 21A and 21B, a circle indicates that the rail pressure is 80 MPa, a triangle indicates that the rail pressure is 128 MPa, and a ◇ indicates that the rail pressure is 180 MPa. Show.

  In order to normalize the fluctuation amount dQt of the main injection to the state shown in FIG. 21B, first, the fluctuation cycle of the fluctuation quantity dQt2 of the main injection, which differs depending on the rail pressure, is a fluctuation of the reference rail pressure, for example, 128 MPa. The interval time Ti at each rail pressure is contracted or extended so as to overlap with the cycle, and then the fluctuation amount dQt2 of the main injection that differs depending on the rail pressure is standardized to a fluctuation pattern at a reference rail pressure, for example, 128 MPa. FIG. 21B shows the fluctuation amount dQt of the main injection standardized in this way.

  Further, although the reference fluctuation pattern as shown in FIG. 21B for the combination of the second cylinder # 2 and the third cylinder # 3 is not particularly shown, this reference fluctuation pattern is also a function of the interval time Ti. It is obtained in advance by experiments. This reference fluctuation pattern is also stored in advance in the ROM 22 together with the reference fluctuation pattern shown in FIG. 21B. Based on these reference fluctuation patterns, the fluctuation amount dQt2 of the main injection in the corresponding fuel injection valve 3 according to the rail pressure. Is calculated.

That is, in order to obtain the fluctuation amount dQt2 of the main injection, first, the interval time Ti is multiplied by the time axis gain FG1 shown in FIG. Using FG1, a reference injection pressure fluctuation amount dQt2 at a reference rail pressure of 128 MPa and a reference main injection amount of 15 (mm ³ ) is obtained from the reference fluctuation pattern shown in FIG. 21B, and then this main injection fluctuation amount dQt2 is obtained. By multiplying the injection amount gain FG3 shown in FIG. 22 (C) corresponding to the main injection amount, the fluctuation amount dQt2 of the main injection corresponding to the main injection amount shown in FIG. 21 (A) is obtained, and then this main injection Is multiplied by the rail pressure gain FG2 shown in FIG. 22B corresponding to the rail pressure to obtain the final main injection variation dQt2 (= The amount of fluctuation of the main injection, which is determined from the level change pattern dQt2 · FG1 · FG2) is required.

  The time axis gain FG1 shown in FIG. 22A is exactly the same as the time axis gain IG1 shown in FIG. 20A. Therefore, the time axis gain FG1 increases as the rail pressure increases. On the other hand, the rail pressure gain FG2 increases as the rail pressure increases as shown in FIG. Since the rail pressure gain FG2 is different for each combination of two cylinders in which each stroke of the engine is separated by 360 degrees in the crank angle, the rail pressure gain FG2 is the first cylinder as shown by a and b in FIG. They are provided for the combination of # 1 and # 4 cylinder # 4 and the combination of # 2 cylinder # 2 and # 3 cylinder # 3, respectively.

  On the other hand, as shown in FIG. 22C, the injection amount gain FG3 increases as the main injection amount increases. Since the injection amount gain FG3 is also different for each combination of two cylinders whose crank angle is 360 degrees apart, the injection amount gain FG3 is the first cylinder as shown by a and b in FIG. 22 (C). They are provided for the combination of # 1 and # 4 cylinder # 4 and the combination of # 2 cylinder # 2 and # 3 cylinder # 3, respectively.

  As shown in FIGS. 18A and 18B, the first cylinder # 1 and the fourth cylinder # 4 are also used when the previous injection is the main injection M and the subsequent injection is the auxiliary fuel injection R. And the reference variation pattern, rail pressure gain FG2, and injection pressure gain FG3 are stored in advance in the ROM 22 for the combination of the second cylinder # 2 and the third cylinder # 3.

  Next, as shown in FIG. 17A, the fuel injection control method will be described by taking as an example the case where the auxiliary fuel injection R is performed before the main injection M and the pilot injection P is performed before the main injection M. To do. In this case, the opening timing and the main injection amount of the needle valve 31 at the time of main injection vary due to the auxiliary fuel injection R and the pilot injection P. At this time, the auxiliary fuel injection R and the pilot injection P are independent of the needle valve 31. It is considered that the valve opening timing and the main injection amount are changed. Therefore, the actual fluctuation amount of the opening timing of the needle valve 31 is the sum of the fluctuation amount Δτ2 of the opening timing of the needle valve 31 caused by the auxiliary fuel injection R and the fluctuation amount Δτ of the opening timing of the needle valve 31 caused by the pilot injection P. Thus, the actual fluctuation amount of the main injection M is the sum of the fluctuation amount dQt2 of the main injection M caused by the auxiliary fuel injection R and the fluctuation amount dQt of the main injection M caused by the pilot injection P.

  Further, when the auxiliary fuel injection R and the pilot injection P are performed, the injection of the auxiliary fuel injection R in one cylinder in the first cylinder # 1 and the fourth cylinder # 4, or the second cylinder # 2 and the third cylinder # 3. The timing and the injection timing of the pilot injection P in the other cylinder are close in time. In this case, if the auxiliary fuel injection R is earlier than the pilot injection P, the valve opening timing of the needle valve 31 at the time of the pilot injection P is changed by the auxiliary fuel injection R and the injection amount is changed. If it is earlier, the opening timing of the needle valve 31 at the time of the auxiliary fuel injection R is changed by the pilot injection P and the injection amount is changed.

  Even in this case, the reference fluctuation pattern of the fluctuation amount of the opening timing of the needle valve 31 in the auxiliary fuel injection R or the pilot injection P and the reference fluctuation pattern of the fluctuation quantity of the injection amount are obtained in advance by experiments, and these reference fluctuation patterns are obtained. The injection amount can be controlled based on the above. In the example shown below, the description of the injection control based on the reference fluctuation pattern of the auxiliary fuel injection R and the pilot injection P is omitted.

  Now, referring to the fuel injection control routine shown in FIGS. 23 to 25, first, in step 100, the total injection amount QT is calculated from the map shown in FIG. 4 (A). Next, at step 101, the main injection amount QM is calculated from the map shown in FIG. Next, at step 102, a pilot injection amount QP is calculated from a map not shown. Next, at step 103, the auxiliary fuel injection amount QR is calculated by subtracting the main injection amount QM and the pilot injection amount QP from the total injection amount QT. Next, at step 104, the main injection start timing θM is calculated from a map not shown. Next, at step 105, the pilot injection timing θP is calculated from a map not shown. Next, at step 106, an interval time TI between the pilot injection P and the main injection M is calculated from the map shown in FIG.

Next, at step 107, the time axis gain IK1 corresponding to the rail pressure is obtained from FIG. Next, at step 108, the corrected interval time Ti is calculated by multiplying the interval time TI by the time axis gain IK1. Next, at step 109, if the reference rail pressure is 80 MPa and the reference pilot injection amount QP is 2 (mm ³ ), that is, the variation amount Δτ of the valve opening timing indicated by a circle in FIG. Then, the reference fluctuation amount of the valve opening timing according to the correction interval time Ti is calculated.

  Next, at step 110, a rail pressure gain IK2 corresponding to the rail pressure is calculated from FIG. 9B. Next, at step 111, the final valve opening timing variation Δτ is calculated by multiplying the reference variation amount Δτ of the valve opening timing calculated at step 109 by the rail pressure gain IK2.

Next, at step 112, the auxiliary fuel injection timing θR is calculated from a map not shown. Next, at step 113, the interval time Ti between the auxiliary fuel injection R and the main injection M is calculated from FIG. Next, at step 114, the time axis gain IG1 corresponding to the rail pressure is obtained from FIG. Next, at step 115, the corrected interval time Ti is calculated by multiplying the time axis gain IG1 by the interval time TI calculated at step 113. Next, at step 116, if the reference rail pressure is 128 MPa and the reference auxiliary fuel injection amount QR is 2 (mm ³ ), that is, the variation amount Δτ2 of the valve opening timing indicated by Δ in FIG. Then, the reference fluctuation amount of the valve opening timing according to the correction interval time Ti is calculated.

  Next, at step 117, a rail pressure gain IG2 corresponding to the rail pressure for the corresponding cylinder combination is calculated from FIG. Next, at step 118, the final valve opening timing variation Δτ2 is calculated by multiplying the reference variation amount Δτ2 of the valve opening timing calculated at step 116 by the rail pressure gain IG2.

Next, at step 119, the time axis gain FK1 corresponding to the rail pressure is calculated from FIG. Next, at step 120, the corrected interval time Ti is calculated by multiplying the time axis gain FK1 by the interval time TI calculated at step 106. Next, at step 121, if the reference rail pressure is 80 MPa, the reference main injection amount QM is 20 (mm ³ ), and the reference pilot injection amount QP is 2 (mm ³ ), that is, Δ in FIG. Is the reference fluctuation amount dQt, the reference fluctuation amount dQt corresponding to the correction interval time Ti is calculated.

  Next, at step 122, a rail pressure gain FK2 corresponding to the rail pressure is calculated from FIG. Next, at step 123, an injection amount gain FK3 corresponding to the main injection amount is calculated from FIG. Next, at step 116, the final fluctuation amount dQt of the main injection is calculated by multiplying the reference fluctuation amount dQt calculated at step 124 by the rail pressure gain FK2 and the injection amount gain FK3.

Next, at step 125, the time axis gain FG1 corresponding to the rail pressure is calculated from FIG. Next, at step 126, the corrected interval time Ti is calculated by multiplying the time axis gain FG1 by the interval time TI calculated at step 113. Next, at step 127, if the reference rail pressure is 128 MPa, the reference main injection amount QM is 15 (mm ³ ), and the reference pilot injection amount QP is 2 (mm ³ ), that is, Δ in FIG. Is the reference fluctuation amount dQt2, the reference fluctuation amount dQt2 corresponding to the correction interval time Ti is calculated.

  Next, at step 128, a rail pressure gain FG2 corresponding to the rail pressure for the corresponding cylinder combination is calculated from FIG. 22B, and then at step 129, according to the main injection amount for the corresponding cylinder combination from FIG. 22C. The injection amount gain FG3 is calculated. Next, at step 130, the final fluctuation amount dQt2 of the main injection is calculated by multiplying the reference fluctuation amount dQt2 calculated at step 127 by the rail pressure gain FG2 and the injection amount gain FG3.

  Next, at step 131, the final valve opening timing fluctuation amount (Δτ + Δτ2) is calculated from the sum of the valve opening timing fluctuation amount Δτ obtained at step 111 and the valve opening timing fluctuation amount Δτ2 obtained at step 118. The final main injection fluctuation amount (dQt + dQt2) is calculated from the sum of the main injection fluctuation amount dQt obtained in step 124 and the main injection fluctuation amount dQt2 obtained in step.

  When the final variation amount (Δτ + Δτ2) of the final needle valve 31 and the final variation amount (dQt + dQt2) of the main injection are obtained in this way, the final variation amount of the valve opening timing of the needle valve 31 is obtained. Based on (Δτ + Δτ2) and the final fluctuation amount (dQt + dQt2) of the main injection, the injection command pulse length for the main injection is corrected so that the actual main injection amount becomes the target value. That is, for example, when the fluctuation amount (dQt + dQt2) is positive, the fluctuation amount (dQt + dQt2) is subtracted from the main injection amount QM calculated in step 101, and the main injection amount obtained by subtracting the injection amount from the corresponding fuel injection valve 3 The injection command pulse length necessary to obtain the quantity (QM-dQt-dQt2) is calculated from the relationship shown in FIG.

  On the other hand, if the fluctuation amount (dQt + dQt2) is negative, the fluctuation amount (dQt + dQt2) is added to the main injection amount QM, and the main injection amount (QM + dQt + dQt2) is obtained by adding the injection amount from the corresponding fuel injection valve 3. The injection command pulse length necessary for this is calculated from the relationship shown in FIG. Subsequently, the final injection command pulse length is obtained by adding the final variation amount (Δτ + Δτ2) of the valve opening timing to the injection command pulse length calculated in this way. In this way, the actual main injection amount is controlled to the target value QT. Next, at step 132, auxiliary fuel injection, pilot injection, and main injection are performed.

  Note that the fluctuation amount dQm (= (ΔQ / Δt) · (Δτ + Δτ2)) of the main injection amount is calculated from ΔQ / Δt shown in FIG. 13 from the calculated fluctuation amount (Δτ + Δτ2) of the final valve opening timing. Then, based on the calculated fluctuation amount dQm of the main injection amount and the final fluctuation amount (dQt + dQt2) of the main injection obtained in step 131, the injection with respect to the main injection is performed so that the actual main injection amount becomes the target value. The command pulse length can also be corrected.

FIG. 1 is an overall view of a compression ignition type internal combustion engine. It is side surface sectional drawing which shows the front-end | tip part of a fuel injection valve. It is a figure which shows an injection pattern. It is a figure which shows maps, such as injection quantity. It is a figure which shows the map of interval time. It is a figure which shows the variation | change_quantity of main injection. It is a figure which shows the variation | change_quantity of main injection. It is a figure which shows the fluctuation amount of the valve opening timing of a needle valve. It is a figure which shows the time-axis gain IK1 and the rail pressure gain IK2. It is a figure which shows the injection amount of an injection command pulse and a needle valve. It is a figure which shows the injection amount of an injection command pulse and a needle valve. It is a figure which shows the relationship between command injection pulse length and injection amount. It is a figure which shows the relationship between (DELTA) Q / (DELTA) t and injection amount. It is a figure which shows the variation | change_quantity of main injection. It is a figure which shows the time-axis gain FK1, the rail pressure gain FK2, and the injection amount gain FK3. It is the figure which took out and showed only the fuel injection valve shown in FIG. 1, and a common rail. It is a figure which shows the injection timing of the auxiliary fuel injection R and the main injection M. It is a figure which shows the injection timing of the auxiliary fuel injection R and the main injection M. It is a figure which shows the fluctuation amount of the valve opening timing of a needle valve. It is a figure which shows the time-axis gain IG1 and the rail pressure gain IG2. It is a figure which shows the variation | change_quantity of main injection. It is a figure which shows the time-axis gain FG1, the rail pressure gain FG2, and the injection amount gain FG3. It is a flowchart which shows fuel-injection control. It is a flowchart which shows fuel-injection control. It is a flowchart which shows fuel-injection control.

Explanation of symbols

2 Combustion chamber 3 Fuel injection valve 12 Fuel supply pipe 13 Common rail 31 Needle valve 32 Suck chamber 34 Nozzle chamber 36 Pressure control chamber

Claims (20)

  The first injection and the compression stroke before and after the compression top dead center, which are provided between the second half of the exhaust stroke before and after the intake top dead center and the first half of the intake stroke, comprising a common rail and a plurality of fuel injection valves connected to the common rail. At least two injections with the second injection performed between the second half and the first half of the expansion stroke are performed during one cycle of the engine in each cylinder, and each stroke of the engine is separated by 360 degrees in crank angle. In the cylinder, one of the first injection in one cylinder and the second injection in the other cylinder is performed before the other, and at this time, the subsequent injection is performed after the previous injection is performed. In the injection control device for an internal combustion engine in which the amount of fluctuation with respect to the target value of the subsequent injection changes depending on the interval time until the time when the rail pressure is a predetermined reference rail pressure, The reference fluctuation amount of the injection after changing along the fluctuation pattern is stored, and the interval time is corrected when calculating the fluctuation amount of the subsequent injection when the rail pressure is not the reference rail pressure from the reference fluctuation amount. The time axis gain for performing is stored, and the rail pressure gain to be multiplied by the reference fluctuation amount when the fluctuation amount of the subsequent injection when the rail pressure is not the reference rail pressure is obtained from the reference fluctuation amount is stored. The fluctuation amount of the subsequent injection is obtained from the reference fluctuation amount using the time axis gain and the rail pressure gain, and the injection amount of the subsequent injection is controlled to the target value using the fluctuation amount. Fuel injection device.   The fuel injection device for an internal combustion engine according to claim 1, wherein the first injection is auxiliary injection, and the second injection is main injection, pilot injection before main injection, or post injection after main injection.   The time axis gain is a function of the rail pressure, and the corrected interval time is obtained by multiplying the interval time by the time axis gain, and the fluctuation amount of the subsequent injection from the reference fluctuation amount using the corrected interval time. The fuel injection device for an internal combustion engine according to claim 1, wherein:   The fuel injection device for an internal combustion engine according to claim 1, wherein the rail pressure gain is a function of the rail pressure.   2. The fuel injection device for an internal combustion engine according to claim 1, wherein the reference fluctuation amount is a fluctuation amount of a valve opening timing of a needle valve of the fuel injection valve.   2. The fuel injection device for an internal combustion engine according to claim 1, wherein the reference fluctuation amount is a fluctuation amount of an injection amount of subsequent injection.   An injection amount gain to be multiplied by the reference variation amount when the variation amount of the injection amount of the subsequent injection when the rail pressure is not the reference rail pressure is obtained from the reference variation amount is stored, and the time axis gain and the rail pressure are stored. The fuel injection device for an internal combustion engine according to claim 6, wherein a fluctuation amount of an injection quantity of a subsequent injection is obtained from the reference fluctuation quantity using the injection quantity gain in addition to the gain.   The fuel injection device for an internal combustion engine according to claim 7, wherein the injection amount gain is a function of an injection amount of subsequent injection.   The reference fluctuation amount is obtained by subtracting the reference fluctuation amount of the needle valve opening timing of the fuel injection valve and the fluctuation amount of the injection amount due to the fluctuation of the needle valve opening timing from the whole fluctuation amount of the injection amount of the subsequent injection. The fuel injection device for an internal combustion engine according to claim 1, comprising a reference fluctuation amount of the injection amount of the subsequent injection obtained by the above.   Both the fluctuation amount of the needle valve opening timing obtained from the reference fluctuation amount of the needle valve opening timing and the fluctuation amount of the injection amount of the subsequent injection obtained from the reference fluctuation amount of the injection amount of the subsequent injection The fuel injection device for an internal combustion engine according to claim 9, wherein the injection command pulse length of the subsequent injection is corrected by the operation.   The relationship between the injection amount and the injection command pulse length is obtained in advance, and the injection amount of the subsequent injection is determined based on the fluctuation amount of the injection amount of the subsequent injection obtained from the reference fluctuation amount of the injection amount of the subsequent injection. The injection command pulse length required to obtain the target value is obtained from the above relationship, and further, the needle valve opening timing obtained from the reference fluctuation amount of the needle valve opening timing to the injection command pulse obtained from the above relationship. The fuel injection device for an internal combustion engine according to claim 10, wherein the fluctuation amount is added.   The amount of change in the injection amount of the subsequent injection based on the change in the valve opening timing of the needle valve is calculated from the amount of change in the valve opening timing of the needle valve obtained from the reference amount of change in the valve opening timing of the needle valve. The injection command for the subsequent injection based on both the fluctuation amount of the injection amount of the subsequent injection based on the fluctuation of the valve opening timing and the fluctuation amount of the injection amount of the subsequent injection obtained from the reference fluctuation amount of the injection amount of the subsequent injection The fuel injection device for an internal combustion engine according to claim 9, wherein the pulse length is corrected.   2. The fuel for an internal combustion engine according to claim 1, wherein there are a plurality of combinations of two cylinders in which each stroke of the engine is separated by 360 degrees in crank angle, and a separate reference fluctuation amount is stored for each combination. Injection device.   The fuel injection device for an internal combustion engine according to claim 13, wherein the rail pressure gain is stored for each of the combinations.   The injection amount gain multiplied by the reference variation amount when the variation amount of the injection amount of the subsequent injection when the rail pressure is not the reference rail pressure is obtained from the reference variation amount is stored for each combination described above. 14. The fuel injection device for an internal combustion engine according to claim 13, wherein a variation amount of the injection amount of the subsequent injection is obtained from the reference variation amount using the injection amount gain in addition to the time axis gain and the rail pressure gain.   A third injection separate from the previous injection is performed prior to the subsequent injection, in addition to the interval time for the previous injection from the previous injection to the subsequent injection, The interval time for the third injection from the third injection to the subsequent injection is determined, and the interval time for the third injection is when the rail pressure is a predetermined reference rail pressure. The reference fluctuation amount for the third injection of the injection after changing along the reference fluctuation pattern with the increase of the reference is stored, and the fluctuation of the subsequent injection due to the third injection when the rail pressure is not the reference rail pressure The time axis gain for the third injection for correcting the interval time when determining the amount from the reference fluctuation amount for the third injection is stored, and the third injection when the rail pressure is not the reference rail pressure is stored. The rail pressure gain for the third injection, which is multiplied by the reference fluctuation amount for the third injection when the fluctuation amount of the subsequent injection due to the third injection is obtained from the reference fluctuation amount for the third injection, is stored. The variation amount of the subsequent injection by the third injection is obtained from the reference variation amount for the third injection using the time axis gain for the third injection and the rail pressure gain for the third injection, and the variation amount of the subsequent injection due to the previous injection The fuel injection device for an internal combustion engine according to claim 1, wherein an injection amount of the subsequent injection is controlled to a target value using a variation amount of the subsequent injection by the third injection.   The previous injection is auxiliary injection performed between the second half of the exhaust stroke before and after the intake top dead center and the first half of the intake stroke, the subsequent injection is the main injection, and the third injection is the pilot injection. Fuel injection device for internal combustion engine.   The reference fluctuation amount for the previous injection is a reference fluctuation amount for the previous injection at the valve opening timing of the needle valve of the fuel injection valve, and the needle based on the previous injection from the entire fluctuation amount of the injection amount of the subsequent injection due to the previous injection The injection amount of the subsequent injection obtained by subtracting the amount of change in the injection amount due to the change in the valve opening timing is a reference variation amount with respect to the previous injection, and the reference variation amount for the third injection is the fuel injection. According to the reference fluctuation amount of the valve opening timing of the needle valve of the valve with respect to the third injection and the fluctuation amount of the needle valve based on the third injection from the whole fluctuation amount of the injection amount of the subsequent injection by the third injection The fuel injection device for an internal combustion engine according to claim 16, comprising a reference fluctuation amount with respect to the third injection of the injection quantity of the subsequent injection obtained by subtracting the fluctuation quantity of the injection quantity.   The total fluctuation amount of the needle valve opening timing is the fluctuation amount of the needle valve opening timing due to the previous injection obtained from the reference fluctuation amount with respect to the previous injection of the needle valve opening timing, and the valve opening timing of the needle valve. And the amount of change in the valve opening timing of the needle valve due to the third injection obtained from the reference amount of change for the third injection, and the total amount of change of the subsequent injection is the amount before the injection amount of the subsequent injection. The amount of change in the injection amount of the subsequent injection due to the previous injection obtained from the reference amount of variation for the injection and the amount of the subsequent injection by the third injection obtained from the reference amount of change of the injection amount of the subsequent injection with respect to the third injection The injection command pulse length of the subsequent injection is corrected by both the total variation amount of the opening timing of the needle valve and the total variation amount of the subsequent injection. 2. A fuel injection device for an internal combustion engine according to 1.   Based on the amount of fluctuation in the needle valve opening timing due to the previous injection obtained from the reference fluctuation amount with respect to the previous injection of the needle valve opening timing, the subsequent injection due to the previous injection based on the variation in the valve opening timing of the needle valve The amount of change in the injection amount is calculated, and the amount of valve opening timing of the needle valve is calculated from the amount of change in the valve opening timing of the needle valve by the third injection obtained from the reference amount of variation of the needle valve opening timing with respect to the third injection. The variation amount of the injection amount of the subsequent injection based on the third injection based on the variation is calculated, the variation amount of the injection amount of the subsequent injection based on the variation of the valve opening timing of the needle valve, and the needle valve Of the subsequent injection by the third injection based on the fluctuation of the valve opening timing of the second injection and the subsequent injection by the previous injection obtained from the reference fluctuation amount of the injection amount of the subsequent injection with respect to the previous injection The amount of fluctuation in the injection amount and the injection amount of the subsequent injection 19. The internal combustion engine according to claim 18, wherein the injection command pulse length of the subsequent injection is corrected from the sum of the fluctuation amount of the injection amount of the subsequent injection by the third injection obtained from the reference fluctuation amount for the third injection. Fuel injection device.

Images