JP2014009656A - Valve full-closed position learning device for valve - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a full-closed position learning device for a valve in which trouble such as fracture of the valve or a stopper can be avoided without fail by surely suppressing a speed of abutting the valve to the stopper within an appropriate range when learning a full-closed position of the valve.SOLUTION: The full-closed position learning device comprises: an EGR valve 14 for controlling a flow rate of EGR gas; an EGR motor 15 for driving the EGR valve 14; an EGR valve aperture sensor 21 for detecting an EGR valve aperture; a stopper 12a to which the EGR valve 14 is abutted at the full-closed position; and a return spring 19 for biasing the EGR valve 14 closer to the full-closed position. Full-closing control is performed to drive the EGR valve 14 in a full-closing direction by the EGR motor 15 and to abut the EGR valve 14 to the stopper 12a and based on the EGR aperture detected during the full-closing control, the full-closed position is learnt. Prior to the full-closing control, electrification stop control is executed to stop electrification to the EGR motor 15 for a first predetermined time.

Description

  The present invention relates to a valve fully closed position learning device that learns a fully closed position of a valve that controls a gas flow rate.

  As a conventional valve fully closed position learning device, for example, a device described in Patent Document 1 below is known. In Patent Document 1, the valve is, for example, an EGR valve provided in an EGR passage of an internal combustion engine. This EGR valve is driven by a motor and controls the flow rate of exhaust gas flowing through the EGR passage. The motor is composed of a DC motor or the like, and drives the EGR valve according to the magnitude and direction of the energized current. The EGR valve is always urged toward the fully closed position by the return spring, and this fully closed position is determined by the EGR valve abutting against a stopper (valve seat).

  When learning the fully closed position of the EGR valve, a predetermined first drive current is supplied to the motor in order to drive the EGR valve to the fully closed position and forcibly abut the stopper. When the opening degree of the EGR valve detected in this state does not substantially change for a predetermined time, the detected opening degree at that time is learned as the fully closed position of the EGR valve. Further, it is determined whether or not the learned fully closed position is within a predetermined range.

  If it is determined that the fully closed position is not within the predetermined range, the EGR valve may not be positioned at the fully closed position due to soot or dust adhesion. In order to abut, the above-described learning and determination of the fully closed position are performed in the same manner in a state where a predetermined third drive current larger than the first drive current is supplied to the motor. Thereafter, until the learned fully closed position falls within a predetermined range, learning and determination of the fully closed position are repeated while further increasing the drive current to the motor stepwise.

JP 2008-196389 A

  As described above, in the conventional learning device, when the fully closed position of the EGR valve is learned, the EGR valve is driven to the fully closed position and forcedly abutted against the stopper. Energization of the first drive current is started immediately. For this reason, when this energization is started from a state in which the EGR valve is relatively far from the stopper, the EGR valve may hit the stopper at an excessive speed due to acceleration by the spring biasing force and the motor driving torque. This may cause damage to EGR valves, stoppers, EGR valve drive mechanisms including motors, and the like.

  In order to avoid such a problem, it is conceivable to start energization of the learning motor on condition that the EGR valve is close to the stopper, that is, the fully closed position. However, due to deterioration of the sensor that detects the opening degree of the EGR valve, temperature characteristics, etc., the opening degree of the EGR valve recognized by the control device may shift to the closing side with respect to the actual opening degree. . In that case, even if the motor is energized under the above conditions, the EGR valve is actually driven from a position farther than the controller recognizes. There is a risk of hitting, and the above-mentioned problems cannot be avoided reliably.

  Further, in this conventional learning device, the stepwise increase of the drive current to the motor and the learning and determination of the fully closed position are repeated until the learned fully closed position falls within a predetermined range. It takes time and the transition to normal control of the EGR valve is delayed, and during that time, the EGR valve is strongly pressed against the stopper by the driving force of the motor, so that the wear and deterioration of the EGR valve easily progress. There are also disadvantages.

  The present invention has been made to solve such problems, and can reliably suppress the pressing speed of the valve to the stopper when learning the fully closed position of the valve within an appropriate range. An object of the present invention is to provide a valve fully closed position learning device that can reliably avoid problems such as breakage of a valve drive mechanism including a valve, a stopper, and a motor.

  In order to achieve this object, a valve fully closed position learning device according to claim 1 of the present application is provided in a gas passage (EGR passage 12 in the embodiment (hereinafter the same in this section)) and gas flowing in the gas passage. A valve (EGR valve 14) for controlling the flow rate of (EGR gas), a motor (EGR motor 15) that operates by energizing current and drives the valve, and the valve opening (detected voltage values EGRLAD, EGR valve) A valve opening degree detecting means (EGR valve opening degree sensor 21) for detecting the opening degree LEGR), a stopper (valve seat 12a) for positioning the valve in a fully closed position by contacting the valve, and a valve in the fully closed position. When a predetermined execution condition for learning the fully closed position of the valve and the biasing member (return spring 19) for biasing is established, the valve is driven by driving the motor. Based on the fully closed control means (ECU2, FIG. 7) for executing the fully closed control for driving the valve in the fully closed direction and forcibly pressing it against the stopper, and the detected opening of the valve detected during the fully closed control, And a fully closed position learning means (ECU2, FIG. 8) for learning a fully closed position (full closed learning value EGRVZR) of the valve. The fully closed control means first energizes the motor prior to the fully closed control. It is characterized in that the energization stop control is stopped for a predetermined time (fourth predetermined time TREF4) (step 55 in FIG. 7).

  This valve is driven by a motor to control the flow rate of the fluid flowing through the gas passage. The valve is urged toward the fully closed position by the urging member and is fully closed by contacting the stopper. Positioned. According to the present invention, when the fully closed position of the valve is learned, the valve is driven in the fully closed direction by driving the motor, and the fully closed control for forcibly pressing the valve against the stopper is executed. Learning of the fully closed position of the valve is performed based on the detected opening degree of the valve detected during the closing control.

  Prior to the fully closed control, the energization stop control is executed to stop energization of the motor for a first predetermined time. Thus, the valve is driven to the fully closed position side by the urging force of the urging member, and is surely positioned near the fully closed position. By starting full-closed control from this state, it is possible to reliably suppress the valve pressing speed against the stopper at this time to an appropriate range that is not excessive, and thus a valve including a valve, a stopper, and a motor. It is possible to reliably avoid problems such as damage to the drive mechanism.

  According to a second aspect of the present invention, in the valve fully closed position learning device according to the first aspect of the invention, is an opening direction deviation in which the detected opening degree of the valve deviates in the opening direction with respect to the actual opening degree of the valve? Open direction deviation determining means (ECU2, steps 26, 30, 31 in FIG. 5) for determining whether or not the full-closed control means determines that an opening direction deviation has occurred, 7 is executed, and when it is determined that the deviation in the opening direction has not occurred, the energization stop control is executed and then the fully closed control is executed (steps 53 to 58 in FIG. 7). ).

  When there is a deviation in the opening direction in which the detected opening of the valve deviates in the opening direction from the actual opening (actual opening), the actual opening of the valve is located at or close to the fully closed position. Therefore, even if the fully closed control is executed as it is, there is no possibility that the valve pressing speed against the stopper becomes excessive. From this point of view, according to the present invention, when it is determined that a deviation in the opening direction has occurred, the fully closed control is executed without executing the energization stop control, so that the above-described problems do not occur. The learning of the fully closed position can be completed in a short time. On the other hand, when it is determined that the deviation in the opening direction has not occurred, the energization stop control is executed and then the fully closed control is executed. Therefore, the above-described operation according to claim 1 can be ensured.

  According to a third aspect of the present invention, in the valve fully closed position learning device according to the second aspect, the target opening degree setting means (ECU2) for setting the target opening degree LCMD of the valve, and the detected opening degree of the valve (EGR valve) Feedback control means (ECU2) that feedback-controls the current (motor duty ratio MDUTY) that energizes the motor so that the opening degree LEGR) becomes the set target opening degree LCMD. In the state where the target opening degree LCMD is set to 0 during the feedback control, when the detected opening degree of the valve is equal to or less than the predetermined value LREF near 0, it is determined that no opening direction deviation has occurred (see FIG. 5 in steps 26 to 28), when the detected opening degree of the valve exceeds the predetermined value LREF even after the second predetermined time TREF2 has elapsed, the opening direction Les and judging a occurs (step 26,30,31 in Figure 5).

  When the energizing current to the motor is feedback controlled so that the detected opening of the valve becomes the set target opening, the detection opening of the valve is not performed when the target opening of the valve is set to 0. As long as there is no deviation in the opening direction, the detected opening of the valve should be below a predetermined value near zero. From such a point of view, according to the present invention, in the above state, when the detected opening of the valve is equal to or less than a predetermined value near 0, it is determined that no opening direction deviation has occurred, and the second predetermined time Even if elapses, if the detected opening of the valve exceeds a predetermined value, it is determined that a deviation in the opening direction has occurred. Thereby, the presence or absence of a deviation in the opening direction of the detected opening of the valve can be accurately determined. Further, based on the determination result, it is possible to appropriately determine whether or not energization stop control prior to full-closed control is possible.

  The invention according to claim 4 is the valve fully-closed position learning device according to any one of claims 1 to 3, wherein the fully-closed control means performs replacement or charging of a battery that is a power source of the valve opening degree detecting means. When this occurs, the fully closed control is executed (steps 1, 4, 5, and 4 in FIG. 3), and the valve fully closed position is learned based on the detected opening of the valve detected during the fully closed control. The battery canceling fully closed position learning means (ECU2, steps 65 and 66 in FIG. 8) is further provided.

  When the battery that is the power source of the valve opening detection means is replaced or charged, the detection opening of the valve opening detection means is likely to change as the voltage characteristics of the battery change. The reliability of the learning value of the fully closed position obtained before that becomes low. According to this configuration, when the battery is replaced or charged, as the learning at the time of battery cancellation, the fully closed control is executed, and based on the detected opening of the valve detected during the fully closed control, Learn the fully closed position. Thereby, the update can be appropriately performed while compensating for the deviation of the learning value at the fully closed position caused by the change in the voltage characteristic of the battery due to the battery cancellation.

  According to a fifth aspect of the present invention, in the valve fully closed position learning device according to any one of the first to fourth aspects, the fully closed control means includes an existing learning value (full closed learning value EGRVZR) of the valve fully closed position. ) Is present, the detected opening of the valve (detected voltage value EGRLAD) detected in a state where the energization of the motor is stopped with the stop of the control of the gas flow rate, and the existing learning value When the deviation degree DV0 is equal to or greater than a predetermined threshold value DREF, the fully closed control is executed (steps 3 to 5 in FIG. 3 and FIG. 6), and based on the detected opening of the valve detected during the fully closed control. Further, it is further characterized by further comprising a fully closed position relearning means (ECU 2, steps 65 and 68 in FIG. 8) for learning again the fully closed position of the valve.

  In the state where the control of the gas flow rate by the valve is stopped and energization of the motor is stopped accordingly, the valve is positioned at or near the fully closed position by the biasing by the biasing member. The detected opening should show a value close to the existing learning value that represents the fully closed position of the valve. From this point of view, according to the present invention, when the degree of deviation between the detected opening of the valve and the existing learning value is equal to or greater than a predetermined threshold value in the above state, the fully closed position at that time is Assuming that there is a deviation from the existing learning value, the fully closed control is executed, and the fully closed position of the valve is relearned based on the detected opening of the valve detected during the fully closed control. Thereby, the learning value can be updated appropriately by performing the relearning of the fully closed position at an appropriate timing while monitoring the deviation of the existing learning value as needed.

  The invention according to claim 6 is the valve fully-closed position learning device according to claim 5, wherein the learning value EGRVZR0 of the fully-closed position is calculated based on the detected opening of the valve detected during the fully-closed control. After the calculation, the fully closed position learning means at the time of battery cancellation sets the current value EGRVZR0 of the calculated learning value as a new learning value EGRVZR instead of the existing learning value (previous value EGRVZRZ), and fully closed position learning The means and the fully closed position relearning means calculate a new learning value EGRVZR using both the existing learning value and the current value EGRVZR0 of the learning value.

  As described above, when learning of the fully closed position is performed due to battery cancellation, the reliability of the learned value learned before that is low. Therefore, in this case, instead of the existing learning value, the current value of the learning value calculated after battery cancellation is set as a new learning value, and only the current value is reflected to update the learning value appropriately. Can be done. On the other hand, when learning of the fully closed position is performed for reasons other than battery cancellation, it is possible to reflect not only the current learning value but also the existing learning value in the new learning value. This is preferable for ensuring reliability and stability. Therefore, in this case, the learning value can be updated appropriately by calculating a new learning value using both the existing learning value and the current value.

It is a figure which shows roughly the internal combustion engine which has an EGR valve to which this invention is applied, and its control apparatus. It is a figure which shows typically the structure of an EGR valve in the state which exists in (a) fully closed position and (B) valve opening position. It is a main flow of the fully closed position learning process of the EGR valve performed by the control device. It is a subroutine which shows the execution determination process of learning at the time of battery cancellation of FIG. It is a subroutine which shows the execution determination process of the learning at the time of feedback control of FIG. It is a subroutine which shows the execution determination process of the learning at the time of EGR stop of FIG. It is a subroutine which shows the fully-closed control process of FIG. It is a subroutine which shows the learning value calculation process of FIG. It is a timing chart which shows the operation example obtained by a fully closed position learning process. It is a timing chart which shows another example of operation obtained by a fully closed position learning process. It is a timing chart which shows another example of operation obtained by a fully closed position learning process.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 shows an internal combustion engine (hereinafter referred to as “engine”) 3 to which the present invention is applied. The engine 3 is a four-cylinder diesel engine mounted on a vehicle (not shown), for example. A combustion chamber 3d is formed between the piston 3b and the cylinder head 3c of each cylinder 3a (only one is shown).

  An intake passage 4 and an exhaust passage 5 are connected to each cylinder 3a, and a fuel injection valve (hereinafter referred to as "injector") 6 is attached so as to face the combustion chamber 3d. The injector 6 directly injects high-pressure fuel into the combustion chamber 3d. The injection amount and the injection timing are controlled by a control signal from the ECU 2 described later.

  The intake passage 4 is provided with an intake shutter 7 constituted by a butterfly valve, and an actuator 8 constituted by, for example, a DC motor is connected to the intake shutter 7. The opening degree of the intake shutter 7 is controlled by controlling the duty ratio ADUTY of the current supplied to the actuator 8 by the ECU 2, thereby adjusting the amount of fresh air drawn into the cylinder 3a.

  In the exhaust passage 5, an exhaust gas purifying device 9 for purifying exhaust gas discharged from the cylinder 3 a and a filter 10 for collecting particulates in the exhaust gas are provided in order from the upstream side.

  The engine 3 is provided with an EGR device 11 for returning a part of the exhaust gas discharged to the exhaust passage 5 to the intake passage 4 as EGR gas. The EGR device 11 includes an EGR passage 12, an EGR valve mechanism 13 provided in the middle of the EGR passage 12, an EGR cooler (not shown), and the like. The EGR passage 12 is connected to the exhaust passage 5 upstream of the exhaust gas purification device 9 and the intake passage 4 downstream of the intake shutter 7.

  The EGR valve mechanism 13 includes an EGR valve 14 disposed in the EGR passage 12 and an EGR motor 15 that drives the EGR valve 14. As shown in FIG. 2, the EGR mechanism 13 further includes a drive gear (not shown) and a link rod 17 provided in a gear box 16 to which the EGR motor 15 is attached. The drive gear has teeth (both not shown) that mesh with the output shaft of the motor 15, and rotates in a decelerated state as the output shaft rotates. Further, the link rod 17 is rotatably connected to the drive gear, and linearly moves in the length direction of the link rod 17 (vertical direction in FIG. 2) as the drive gear rotates.

  The EGR motor 15 is a DC motor, and rotates in the forward direction or the reverse direction depending on the direction of the energized current. The EGR valve 14 is configured by a poppet valve, and one end portion of the valve shaft 14a is connected to the link rod 17 via a connecting pin 17a, and the other end portion has a valve body 14b. Opens and closes the EGR passage 12 by contacting and disconnecting the valve seat 12 a formed in the EGR passage 12.

  The valve shaft 14a of the EGR valve 14 is integrally provided with a flange-shaped spring receiving portion 14c. The EGR valve 14 is closed between the spring receiving portion 14c and the non-moving spring receiving plate 18. A coil spring-like return spring 19 that is always urged in a direction (upward in FIG. 2) is disposed.

  In the EGR valve mechanism 13 having the above configuration, the lift amount (hereinafter referred to as “EGR valve opening”) LEGR of the EGR valve 14 is the duty ratio (hereinafter referred to as “motor duty ratio”) of the current supplied to the EGR motor 15 MDUTY. Is controlled by the ECU 2, whereby the amount of EGR that recirculates to the intake passage 4 is adjusted.

  Specifically, when the motor duty ratio MDUTY = 0 and the energization of the EGR motor 15 is stopped, the EGR valve 14 is positioned in the upper part of FIG. The valve seat 12a is held in the fully closed position while being seated ((a) in the figure).

  From this state, when a current in a predetermined direction is supplied to the EGR motor 15, the rotation of the EGR motor 15 is converted into a straight movement of the link rod 17 via the drive gear, and the link rod 17 and the EGR valve connected thereto When 14 moves against the urging force of the return spring 19 in FIG. 2, the valve body 14b is separated from the valve seat 12a, and the EGR valve 14 is opened ((b) in FIG. 2). The EGR valve opening degree LEGR at this time is determined by the balance between the driving force of the EGR motor 15 and the urging force of the return spring 19 according to the motor duty ratio MDUTY.

  On the other hand, when a current in the reverse direction is applied to the EGR motor 15, the driving force of the EGR motor 15 in addition to the biasing force of the return spring 19 acts on the EGR valve 14 upward in FIG. The valve 14 is located at the fully closed position in FIG. 5A, and the valve body 12b is forcibly pressed against the valve seat 12a. Hereinafter, the motor duty ratio MDUTY when the current is supplied in the reverse direction is referred to as a “negative duty ratio”.

Further, the EGR valve 14 is provided with an EGR valve opening degree sensor 21. The EGR valve opening degree sensor 21 is configured by a hall sensor or the like, and outputs a detection signal having a voltage (hereinafter referred to as “detection voltage value”) EGRLAD corresponding to the lift amount (position) of the EGR valve 14 to the ECU 2. . The ECU 2 calculates the EGR valve opening degree LEGR by the following equation (1) based on the detected voltage value EGRLAD.
LEGR = (EGRLAD−VOFFSET) · KVL (1)
Here, VOFFSET is an offset voltage to be output in the fully closed position (LEGR = 0) of the EGR valve 14, and KVL is a conversion coefficient. In the fully closed position learning described later, the offset voltage VOFFSET is learned in a state where the EGR valve 14 is controlled to the fully closed position.

  A crank angle sensor 22 is provided on the crankshaft 3 e of the engine 3. The crank angle sensor 22 outputs a CRK signal, which is a pulse signal, to the ECU 2 at every predetermined crank angle (for example, 30 degrees) with the rotation of the crankshaft 3e. The ECU 2 calculates the engine speed (hereinafter referred to as “engine speed”) NE of the engine 3 based on the CRK signal.

  The ECU 2 also receives a detection signal from the engine water temperature sensor 23 indicating the temperature TW of the coolant flowing through the main body of the engine 3 (hereinafter referred to as “engine water temperature”) TW from the accelerator opening sensor 24. The voltage of a battery (not shown) serving as a power source for the EGR motor 15 and the EGR valve opening sensor 21 is detected from the voltage sensor 25 by a detection signal representing an operation amount (hereinafter referred to as “accelerator opening”) AP. Detection signals representing VB (hereinafter referred to as “battery voltage”) are respectively input.

  The ECU 2 is composed of a microcomputer including a CPU, a RAM, a ROM, an I / O interface (all not shown), and the like. The ECU 2 determines the operating state of the engine 3 according to the detection signals of the various sensors 21 to 25 described above, and controls the fuel injection control to the engine 3 and the intake shutter 7 according to the determined operating state. Various control processes such as intake air amount control and EGR control via the EGR valve 14 are executed.

  In this EGR control, the ECU 2 determines whether to execute or stop the EGR according to the determined operating state. When the EGR is stopped, the motor duty ratio MDUTY is set to 0 and the EGR valve 14 is fully closed. Control to position. When executing EGR, the target opening degree LCMD of the EGR valve 14 is set according to the detected accelerator opening degree AP and the engine speed NE, and the detection output from the EGR valve opening degree sensor 21 is set. Based on the voltage value EGRLAD, the EGR valve opening degree LEGR is calculated by equation (1). Then, the motor duty ratio MDUTY is feedback-controlled so that the EGR valve opening degree LEGR becomes the target opening degree LCMD.

  Furthermore, the ECU 2 executes learning control for learning the fully closed position of the EGR valve 14, more specifically, the offset voltage VOFFSET in the equation (1). In the present embodiment, the ECU 2 includes a fully closed control means, a fully closed position learning means, an opening direction deviation determining means, a target opening setting means, a feedback control means, a battery closing fully closed position learning means, and a fully closed position relearning. Corresponds to means.

Hereinafter, the above-described learning control of the fully closed position of the EGR valve 14 will be described in detail. In the present embodiment, the fully closed position learning is configured by learning the following three patterns according to the cause of the execution, and is executed when each execution condition is satisfied.
A. B. Learning when canceling battery Learning during feedback control Learning when EGR stops

  The battery-learning learning is for compensating for a shift in the fully closed position (offset voltage VOFFSET) due to a change in the voltage characteristics of the battery when battery cancellation (battery replacement or charging) is performed. The learning at the time of feedback control is for compensating for the shift of the fully closed position due to the temperature change (rise) of the EGR valve opening sensor 21 during the execution of EGR during the above-described EGR feedback control. Further, the learning at the time of EGR stop is for compensating for the shift of the fully closed position due to the temperature change (decrease) of the EGR valve opening sensor 21 while the EGR is stopped.

  Next, the fully closed position learning control process executed by the ECU 2 will be described in detail with reference to FIGS. These processes are repeatedly executed every predetermined time.

  In the main flow shown in FIG. 3, first, in step 1 (illustrated as “S1”, the same applies hereinafter) to step 3, the above-described learning during battery cancellation (B / C), learning during feedback (F / B) control, or EGR stop. It is determined whether or not the time learning execution condition is satisfied. As will be described later, when it is determined that the execution condition is satisfied in these determination processes and learning is permitted, the learning permission flag F_LRNB / C, F_LRNF / B or FLRNE / S is set to “1”. .

  Next, it is determined whether or not these learning permission flags F_LRNB / C, F_LRNF / B or FLRNE / S are “1” (step 4). If the answer is NO and the execution condition is not satisfied for the fully closed position learning of any pattern, the present process is terminated as it is.

  If the answer to step 4 is YES and the execution condition for the fully closed position learning of any pattern is satisfied, the fully closed control process of the EGR valve 14 is executed (step 5). This full-closed control process performs full-closed control for driving the EGR valve 14 in the full-close direction by driving the EGR motor 15 and forcibly pressing it against the stopper for learning the full-closed position. Next, a learning value calculation process is executed (step 6), and this process ends. In this learning value calculation process, a learning value (hereinafter referred to as “full closing learning value”) EGRVZR of the fully closed position is calculated according to the detected voltage value EGRLAD output from the EGR valve opening sensor 21 during the fully closing control. Is.

  Hereafter, each process of the main flow mentioned above is demonstrated in detail in order. FIG. 4 is a subroutine for execution determination processing for learning at the time of battery cancellation executed in step 1 of FIG. In this process, first, in step 11, it is determined whether or not the battery cancel flag F_B / C is “1”. If the answer is YES and the battery is replaced or charged, it is determined whether or not the learning completion flag F_LRNB / CE is “1” (step 12). If the answer is NO and learning at the time of battery cancellation is not yet executed, it is determined that the execution condition is satisfied, the learning permission flag F_LRNB / C is set to “1” (step 13), and this processing is ended. To do.

  When the answer to step 11 is NO and the battery is not replaced or charged, or when the answer to step 12 is YES and learning at the time of battery cancellation has already been executed, learning at the time of battery cancellation is not executed. It is assumed that the learning permission flag F_LRNB / C is set to “0” (step 14), and this process is terminated. As is clear from the above, learning at the time of battery cancellation is executed only once immediately after the battery is exchanged or charged, for example, when the ignition switch is turned on.

  FIG. 5 is a subroutine for execution determination processing for learning during feedback control executed in step 2 of FIG. In this process, first, in step 21, it is determined whether or not a learning completion flag F_LRNF / BE during feedback control is “1”. The learning completion flag F_LRNF / BE is set to “1” when the feedback control learning is completed, and is reset to “0” when, for example, EGR feedback control is started.

  If the answer to step 21 is YES and learning at the time of feedback control is already completed, it is determined that the execution condition is not satisfied, and the learning permission flag F_LRNF / B is set to “0” (step 22). The process ends. If the answer to step 21 is NO, it is determined whether or not the battery voltage VB is equal to or higher than a predetermined voltage VREF (step 23). When this answer is NO, the detected voltage value EGRLAD of the EGR valve opening sensor 21 may not be stable, so that the execution condition is not satisfied and the step 22 is executed.

  If the answer to step 23 is YES, it is determined whether the feedback control flag F_EGRF / B is “1” and whether the target opening degree LCMD of the EGR valve 14 is 0 (step). 24, 25). If the answer to step 24 is NO and EGR feedback control is not being performed, or if the answer to step 25 is NO and the target opening degree LCMD of the EGR valve 14 is not 0, it is determined that the execution condition is not satisfied, 22 is executed.

  On the other hand, when both of the answers to steps 24 and 25 are YES, that is, when EGR feedback control is being performed and the target opening degree LCMD = 0 of the EGR valve 14, the EGR valve opening degree LEGR at that time is close to zero. It is determined whether or not the value is equal to or less than a predetermined value LREF (eg, 0.05 mm) (step 26). If the answer is YES, it is determined whether or not the value TM_F / B of the timer that measures the duration of the state of LEGR ≦ LREF is equal to or greater than the first time TREF1 (for example, 10 ms) (step 27). When this answer is NO, it is determined that the execution condition is not satisfied, and the step 22 is executed.

  On the other hand, when the answer to step 27 is YES, that is, in the state where the target opening degree LCMD = 0 of the EGR valve 14 is set during EGR feedback control, the state of LEGR ≦ LREF is not less than the first time TREF1, When it continues, it is determined that the opening direction deviation of the EGR valve opening degree LEGR has not occurred, the opening direction deviation flag F_DOPEN is set to “0” (step 28), and the execution condition for learning during feedback control is satisfied. As a result, the learning permission flag F_LRNF / B is set to “1” (step 29), and this process is terminated.

  If the answer to step 26 is NO and EGR valve opening LEGR> predetermined value LREF, it is determined whether or not a timer value TM_DOPEN for measuring the duration of the state is equal to or greater than a second time TREF2 (for example, 150 ms). (Step 30). When this answer is NO, it is determined that the execution condition is not satisfied, and the process proceeds to step 22.

  On the other hand, if the answer to step 30 is YES, that is, during EGR feedback control, the EGR valve opens even if the second predetermined time TREF2 elapses after the target opening degree LCMD of the EGR valve 14 is set to 0. When the degree LEGR exceeds the predetermined value LREF, it is determined that the opening direction deviation of the EGR valve opening degree LEGR has occurred, the opening direction deviation flag F_DOPEN is set to “1” (step 31), and feedback is performed. Assuming that the execution condition of the control time learning is satisfied, the process proceeds to step 29, where the learning permission flag F_LRNF / B is set to “1”, and this process is terminated.

  FIG. 6 is a subroutine for execution determination processing for EGR stop learning executed in step 3 of FIG. In this process, first, in step 41, it is determined whether or not a learning completion flag F_LRNE / SE when EGR is stopped is “1”. This learning completion flag F_LRNE / SE is set to “1” when the learning at the time of EGR stop is completed, and is reset to “0” when the stop of the EGR is started, for example.

  If the answer to step 41 is YES and learning at the time of EGR stop has already been completed, it is determined that the execution condition is not satisfied, and the learning permission flag F_LRNE / S is set to “0” (step 42). The process ends. When the answer to step 41 is NO, as in step 23 of FIG. 5, it is determined whether or not the battery voltage VB is equal to or higher than a predetermined voltage VREF (step 43). When this answer is NO, it is determined that the execution condition is not satisfied, and the step 42 is executed.

  When the answer to step 43 is YES, it is determined whether or not the engine water temperature TW is equal to or higher than a predetermined temperature TWREF (for example, 50 ° C.) (step 44). When this answer is NO, it is determined that the execution condition is not satisfied, and the step 42 is executed.

  If the answer to step 44 is YES, it is determined whether the EGR execution flag F_EGR is “1” and whether the motor duty ratio MDUTY is 0 (steps 45 and 46), respectively. If the answer to step 45 is YES and the EGR is being executed, or if the answer to step 46 is NO and the motor duty ratio MDUTY is not 0, the execution condition is not satisfied and the step 42 is executed. .

  On the other hand, when the answer to step 45 is NO and the answer to step 46 is YES, that is, when the EGR is stopped and the motor duty ratio MDUTY = 0, the detected voltage of the EGR valve opening sensor 21 at that time The absolute value of the difference between the value EGRLAD and the fully closed learning value EGRVZR is calculated as the degree of divergence DV0 between them (step 47). Next, it is determined whether or not the calculated deviation degree DV0 is equal to or greater than a predetermined threshold value DREF (step 48). When this answer is NO, it is determined that the execution condition is not satisfied, and the step 42 is executed.

  If the answer to step 48 is YES, it is determined whether or not the value TM_E / S of a timer that counts the duration of the state where DV0 ≧ DREF is equal to or greater than a third time TREF1 (for example, 60 ms) (step 49). . When this answer is NO, it is determined that the execution condition is not satisfied, and the step 42 is executed.

  On the other hand, when the answer to step 49 is YES, that is, while the EGR is stopped, the state in which the deviation degree DV0 is equal to or greater than the threshold value DREF continues for the third time TREF3 or more when the motor duty ratio MDUTY = 0. In some cases, assuming that the execution condition of learning at EGR stop is satisfied, the learning permission flag F_LRNE / S is set to “1” (step 50), and this process is terminated.

  FIG. 7 is a subroutine for the fully closed control process executed in step 5 of FIG. As described above, in this full-closed control process, as a result of the determination by the determination processes of FIGS. 4 to 6 described so far, the execution condition of learning at battery cancellation, learning at feedback control or learning at EGR stop is satisfied. When it is executed.

  In this process, first, in step 51, it is determined whether or not the fully closed control end flag F_CONVSE is “1”. As will be described later, this full-closed control end flag F_CONVSE is set to “1” to end the full-closed control when a predetermined time has elapsed after the start of the full-closed control. Accordingly, when the answer to step 51 is YES, the process is terminated as it is.

  When the answer to step 51 is NO, it is determined whether the learning permission flag F_LRNF / B at the time of feedback control is “1” and whether the opening direction deviation flag F_DOPEN is “1”. (Steps 52 and 53). When the answer to step 52 is YES and the answer to step 53 is NO, that is, it is determined that the current fully closed position learning is learning during feedback control, and that there is no deviation in the opening direction of the EGR valve opening LEGR. If it is, the energization stop control for stopping energization of the motor is executed in steps 54 to 56 prior to the fully-closed control after step 58 described later.

  First, in step 54, it is determined whether or not an energization stop control completion flag F_MDUTY0E is “1”. If the answer is NO and energization stop control is not completed, the motor duty ratio MDUTY is set to 0 (step 55), and energization to the motor is stopped. Next, it is determined whether or not a timer value TM_MDUTY0 that measures the elapsed time from the start of energization stop is equal to or longer than a fourth predetermined time TREF4 (for example, 10 ms) (step 56). When this answer is NO, the present process is finished as it is, and the energization stop control is continued.

  On the other hand, when the answer to step 56 is YES, that is, when the fourth predetermined time TREF4 has elapsed after the start of the energization stop of the EGR motor 15, the energization stop control completion flag F_MDUTY0E is assumed to end the energization stop control. Is set to "1" (step 57), and this process is terminated.

  When step 57 is executed, the answer to step 54 becomes YES. In this case, the process proceeds to step 58 to start full-closed control. If the answer to step 53 is YES and the opening direction deviation flag F_DOPEN = 1, or if the answer to step 52 is NO and the learning permission flag F_LRNF / B = 0, the process directly proceeds to step 58.

  As described above, when the execution condition of the fully closed position learning is established, when the fully closed position learning is feedback control learning and it is determined that the opening direction deviation of the EGR valve opening LEGR has not occurred, After the energization stop control is executed, the fully closed control is started. In contrast, when the fully closed position learning is feedback control learning and it is determined that a deviation in the opening direction has occurred, or when the fully closed position learning is battery cancel learning or EGR stop learning, energization is stopped. Fully closed control is immediately started without executing control.

  In this fully closed control, first, in step 58, the motor duty ratio MDUTY is set to a negative predetermined value DUTYN (for example, -15%). As a result, the EGR valve 14 is driven in the fully closed direction by the EGR motor 15, and the driving force acts on the EGR valve 14 in addition to the urging force of the return spring 19, whereby the valve body 14 b is moved to the valve seat 12 a. Forced to be pressed.

  Next, it is determined whether or not a timer value TM_MDUTYN that measures the elapsed time from the start of the fully closed control is equal to or greater than a fifth predetermined time TREF5 (step 59). When this answer is NO, this process is terminated as it is, and the fully closed control is continued.

  On the other hand, when the answer to step 59 is YES, that is, when the fifth predetermined time TREF5 has elapsed after the start of the fully closed control, the fully closed control end flag F_CONVSE is set to “1”. (Step 60), and this process is terminated. This is to avoid problems such as wear and deterioration caused by the valve body 14b of the EGR valve 14 being pressed against the valve seat 12a for a long time by the fully closed control.

  FIG. 8 is a subroutine of learning value calculation processing executed in step 6 of FIG. In this process, first, in step 61, the detected voltage value EGRLAD of the EGR valve opening sensor 21 detected during the fully closed control is sampled. Specifically, predetermined N detection voltage values EGRLAD (EGRLAD1 to EGRLADn) are stocked by sequentially storing the detection voltage value EGRLAD in a predetermined storage unit of the buffer every execution cycle of this process.

  Next, a difference between the maximum value MAXEGRLAD and the minimum value MINEGRLAD among the N detected voltage values EGRLAD stored is calculated as a voltage value change width DLAD (step 62). Next, it is determined whether or not the calculated voltage value change width DLAD is equal to or smaller than a predetermined value DLADREF (step 63).

  When the answer to step 63 is NO, it is determined that the detected voltage value EGRLAD detected is not stable, and this process is ended as it is. In this case, N sampling states are maintained by excluding the earliest detected voltage value EGRLAD1 and adding the detected voltage value EGRLAD as the latest EGRLADn in the sampling of the step 61 after the next time. The

  When the answer to step 63 is YES and the voltage value variation width DLAD ≦ the predetermined value DLADREF, the detected voltage value EGRLAD is assumed to be stable, and the latest voltage detected value EGRLADn at that time is set to the fully closed learning value. The current value EGRVZR0 is calculated (step 64).

  Next, it is determined whether or not the learning permission flag F_LRNB / C at the time of battery cancellation is “1” (step 65). If the answer to this question is YES and the current fully closed position learning is the battery cancellation learning, the current value EGRVZR0 calculated in step 64 is directly calculated as a new fully closed learning value EGRVZR (step 66). Next, the learning completion flag F_LRNB / CE at the time of battery cancellation is set to “1” (step 67), and this process is terminated.

  On the other hand, if the answer to step 65 is NO and the current fully closed position learning is learning during feedback control or EGR stop learning, the simple average value of the previous value EGRVZRZ and the current value EGRVZR0 of the fully closed learning value is newly set. The fully closed learning value EGRVZR is calculated (step 68).

  Next, it is determined whether or not the learning permission flag F_LRNF / B during feedback control is “1” (step 69). If the answer is YES and the current fully closed position learning is the feedback control learning, the learning completion flag F_LRNF / BE is set to “1” (step 70). On the other hand, if the answer to step 69 is NO and the fully closed position learning is EGR stop time learning, the learning completion flag F_LRNE / SE is set to “1” (step 71), and this process ends.

  9 to 11 show operation examples obtained by the fully closed position learning process described so far. FIG. 9 is an example of a case where no deviation in the opening direction of the EGR valve opening degree LEGR occurs during EGR feedback control.

  In this example, at time t1, feedback control is started from the EGR stop state, and the learning completion flag F_LRNF / BE during feedback control is reset to “0”. During this feedback control, the motor duty ratio MDUTY is feedback controlled so that the EGR valve opening degree LEGR becomes the target opening degree LCMD.

  Thereafter, as the target opening degree LCMD is set to 0, the EGR valve opening degree LEGR decreases and becomes equal to or less than a predetermined value LREF in the vicinity of 0 (time point t2), and this state is the first predetermined time TREF1, When continuing (time t3), the opening direction deviation flag F_DOPEN is set to “0” on the assumption that the execution condition of the feedback control learning is satisfied in the state where the opening direction deviation of the EGR valve opening LEGR has not occurred. (Step 28 in FIG. 5), the learning permission flag F_LRNF / B is set to “1” (step 29 in FIG. 5), and fully closed position learning is started.

  In this case, since the opening direction deviation flag F_DOPEN = 0 and the answer to step 53 in FIG. 7 is NO, the motor duty ratio MDUTY is set to 0 (step 55 in FIG. 7) prior to the fully closed control. Thus, the energization stop control of the EGR motor 15 is started. As a result, the EGR valve 14 is reliably positioned in the vicinity of the fully closed position by the urging of the return spring 19.

  Thereafter, when the fourth predetermined time TREF4 elapses (time point t4), the energization stop control is finished, and the motor duty ratio MDUTY is set to a negative predetermined value DUTYN (step 58 in FIG. 7) to be fully closed. Control begins. Thus, the EGR valve 14 is driven to the fully closed position side by the EGR motor 15, whereby the valve body 14b is forcibly pressed against the valve seat 12a.

  Then, based on the detected voltage value EGRLAD detected by the EGR valve opening sensor 21 sampled during the fully closed control, when the predetermined calculation condition described above is satisfied by the learning calculation process of FIG. A current value EGRVZR0 of the closed learning value is calculated (step 64 in FIG. 8), and a simple average value of the current value EGRVZR0 and the previous value EGRVZRZ is calculated as a new fully closed learning value EGRVZR (FIG. 8). Step 68). Then, the learning completion flag F_LRNF / BE is set to “1” (step 70 in FIG. 8), and the learning permission flag F_LRNF / B is reset to “0” (step 22 in FIG. 5). Is completed.

  FIG. 10 shows an operation example in the case where a deviation in the opening direction of the EGR valve opening degree LEGR has occurred during EGR feedback control. In this example, EGR feedback control is started at time t6, and the target opening degree LCMD is set to 0 during this feedback control (time t7). Thereafter, when the state in which the EGR valve opening degree LEGR is larger than the predetermined value LREF continues for the second predetermined time TREF2 (time point t8), feedback control is performed in a state in which the opening direction deviation of the EGR valve opening degree LEGR has occurred. Assuming that the time learning execution condition is satisfied, the opening direction deviation flag F_DOPEN is set to “1” (step 31 in FIG. 5), and the learning permission flag F_LRNF / B is set to “1” (step in FIG. 5). 29) The fully closed position learning is started.

  In this case, since the opening direction deviation flag F_DOPEN = 1 and the answer to step 53 in FIG. 7 is YES, the energization stop control of the EGR motor 15 is not executed, and the fully closed control is immediately started. The subsequent operation is the same as in the case of FIG. 9, and when a predetermined calculation condition is satisfied based on the detected voltage value EGRLAD of the EGR valve opening sensor 21 sampled during the fully closed control (time point t9), A simple average value of the current value EGRVZR0 of the fully closed learning value and the previous value EGRVZRZ is calculated as a new fully closed learning value EGRVZR. Further, the learning completion flag F_LRNF / BE is set to “1” and the learning permission flag F_LRNF / B is set to “0”, and the fully closed position learning is completed.

  FIG. 11 shows an operation example in the case of EGR stop learning. In this example, while the EGR with the motor duty ratio MDUTY set to 0 is stopped, the detected voltage value EGRLAD of the EGR valve opening sensor 21 starts to deviate from the fully closed learning value EGRVZR (time point t10), and then the difference between the two A certain degree of divergence DV0 gradually increases. Along with this, the EGR valve opening degree LEGR similarly changes with respect to the target opening degree LCMD (= 0).

  When the deviation degree DV0 becomes equal to or greater than the predetermined threshold value DREF (time t11) and the state continues for the third predetermined time TREF3 (time t12), the answer to step 49 in FIG. 6 becomes YES. As a result, the learning permission flag F_LRNE / S is set to “1” (step 50 in FIG. 6), assuming that the execution condition for learning at EGR stop is satisfied. At the same time, by setting the motor duty ratio MDUTY to a negative predetermined value DUTYN (step 58 in FIG. 7), the fully closed control is immediately started.

  The subsequent operation is the same as in the case of FIGS. 9 and 10, and when a predetermined calculation condition is established based on the detected voltage value EGRLAD of the EGR valve opening sensor 21 sampled during the fully closed control (time point) t13) A simple average value of the current value EGRVZR0 of the fully closed learning value and the previous value EGRVZRZ is calculated as a new fully closed learning value EGRVZR. Further, the learning completion flag F_LRNE / SE is set to “1” (step 71 in FIG. 8), the learning permission flag F_LRNE / S is reset to “0” (step 42 in FIG. 6), and the fully closed position learning is performed. Is completed.

  Although not shown, when battery cancellation is performed, for example, immediately after the ignition switch is turned on, the learning permission flag F_LRNB / C is set to “1” assuming that the execution condition of learning at the time of battery cancellation is satisfied. (Step 13 in FIG. 4), the fully closed control is immediately started (step 58 in FIG. 7). Further, the current value EGRVZR0 of the fully closed learning value is calculated based on the detected voltage value EGRLAD of the EGR valve opening sensor 21 sampled during the fully closed control, and updated as a new fully closed learning value EGRVZR as it is. (Step 66 in FIG. 8).

  As described above, according to the present embodiment, when the fully closed position learning of the EGR valve 14 is performed, the energization to the EGR motor 15 is stopped for the fourth predetermined time TREF4 by the energization stop control prior to the fully closed control. . As a result, the EGR valve 14 is driven to the fully closed position side by the urging of the return spring 19 and is surely positioned in the vicinity of the fully closed position. Then, by starting the fully closed control from this state, the EGR valve 14 is driven in the fully closed direction by the EGR motor 15 and the valve body 14b is pressed against the valve seat 12a. The contact speed can be surely suppressed to an appropriate range that is not excessive, and therefore problems such as breakage of the EGR valve 14, the valve seat 12 a, and the EGR valve mechanism 13 including the EGR motor 15 can be reliably avoided.

  Further, when it is determined that the opening direction deviation of the EGR valve opening LEGR has occurred, the fully closed control is executed without executing the energization stop control. Learning can be completed in a short time. Further, during the EGR feedback control, when the target opening degree LCMD is set to 0 and the EGR valve opening degree LEGR exceeds the predetermined value LREF in the vicinity of 0 even if the second predetermined time TREF2 has elapsed. In addition, it is determined that a deviation in the opening direction has occurred. Therefore, the presence / absence of the opening direction deviation can be accurately determined, and whether or not the energization stop control prior to the fully closed control can be appropriately determined based on the determination result.

  Further, when battery replacement or charging is performed, learning at the time of battery cancellation is executed as learning of the fully closed position of the EGR valve 14, and therefore the fully closed learning value resulting from the change in the voltage characteristics of the battery accompanying the battery cancellation The update can be appropriately performed while compensating for the shift of the EGRVZR.

  Further, in a state where energization to the EGR motor 15 is stopped along with the EGR stop, the degree of deviation DLAD between the detected voltage value EGRLAD of the EGR valve opening sensor 21 and the existing fully closed learning value EGRVZR is predetermined. When the state of being equal to or greater than the threshold value DREF continues for the third predetermined time TREF3 or longer, learning at the time of EGR stop is executed as learning of the fully closed position of the EGR valve 14. Accordingly, the fully closed learning value EGRVZR can be appropriately updated by performing the relearning of the fully closed position at an appropriate timing while monitoring the deviation of the existing fully closed learning value EGRVZR as needed.

  Further, when learning at the time of battery cancellation is executed, the current value EGRVZR0 of the fully closed learning value obtained thereby is updated as it is as the fully closed learning value EGRVZR, while learning at feedback control or learning at EGR stop is performed. When executed, the average value of the current value EGRVZR0 of the fully closed learning value and the previous value EGRVZRZ is updated as the fully closed learning value EGRVZR. In this way, by properly using the final calculation method of the fully closed learning value EGRVZR according to the cause of the fully closed position learning, the fully closed learning value EGRVZR can be updated more appropriately.

  In addition, this invention can be implemented in various aspects, without being limited to the described embodiment. For example, in the embodiment, as a parameter for determining the deviation of the detected opening of the EGR valve 14, the execution determination of feedback control learning uses the EGR valve opening LEGR, and the EGR valve stop learning execution determination Although the detection voltage value EGRVLAD of the EGR valve opening degree sensor 21 is used, this relationship may be reversed.

  In addition, in the execution determination of learning when the EGR valve is stopped, the absolute value DV0 of the difference between the detected voltage value EGRVLAD and the fully closed learning value EGRVZR is used, but not limited to this, the ratio between the two is used. It may be used. Furthermore, in learning at the time of feedback control and learning at the time of EGR valve stop, when calculating the final fully closed learning value EGRVZR using the current value EGRVZR and the previous value EGRVZRZ, both simple average values are used. Of course, a weighted average value may be used.

  In the embodiment, the target for learning the fully closed position of the EGR valve 14 is the offset voltage VOFFSET that should be output at the fully closed position (LEGR = 0) of the EGR valve 14. A correction value for correcting the EGR valve opening EGR to 0 at the position may be a learning target.

  Further, the embodiment is an example in which the present invention is applied to the EGR valve 14 of the engine 3. However, the present invention is not limited to this as long as it satisfies the requirements such as being driven by a motor and contacting the stopper in the fully closed position. The present invention may be applied to a simple valve such as the intake shutter 7 of the embodiment, or may be applied to a valve used for other than the internal combustion engine. In the embodiment, the valve seat 12a formed in the EGR passage 12 is also used as a stopper, but it is needless to say that a separate stopper may be provided. In addition, it is possible to appropriately change the detailed configuration within the scope of the gist of the present invention.

2 ECU (fully closed control means, fully closed position learning means, open direction deviation judging means, target opening setting means, feedback control means, fully closed position learning means upon battery cancellation, fully closed position relearning means)
12 EGR passage (gas passage)
12a Valve seat (stopper)
14 EGR valve (valve)
15 EGR motor (motor)
19 Return spring (biasing member)
21 EGR valve opening sensor (valve opening detecting means)
EGRLAD detection voltage (valve opening)
LEGR EGR valve opening (valve opening)
Target opening of LCMD EGR valve MDUTY Motor duty ratio (current applied to motor)
EGRVZR fully closed learning value (valve fully closed position, learning value)
EGRVZR0 Current value of fully closed learning value (current value of learning value)
EGRVZRZ Last closed learning value (existing learning value)
LREF predetermined value DV0 Deviation degree DREF threshold value TREF2 second predetermined time TREF4 fourth predetermined time (first predetermined time)

Claims (6)

A valve provided in the gas passage for controlling the flow rate of the gas flowing through the gas passage;
A motor that operates by energization of current and drives the valve;
A valve opening degree detecting means for detecting the opening degree of the valve;
A stopper that positions the valve in a fully closed position by contacting the valve;
A biasing member that biases the valve toward the fully closed position;
Full-closed control for driving the valve in the fully-closed direction and forcibly pressing the stopper against the stopper by driving the motor when a predetermined execution condition for learning the fully-closed position of the valve is satisfied. A fully-closed control means for executing
A fully closed position learning means for learning a fully closed position of the valve based on a detected opening of the valve detected during the fully closed control,
The valve fully-closed position learning device is characterized in that the fully-closed control means executes energization stop control for stopping energization of the motor for a first predetermined time prior to the fully-closed control. Further comprising an opening direction deviation determining means for determining whether or not an opening direction deviation in which the detected opening degree of the valve is shifted in the opening direction with respect to the actual opening degree of the valve has occurred,
When it is determined that the opening direction deviation has occurred, the full closing control means executes the full closing control without executing the energization stop control, and the opening direction deviation has not occurred. 2. The valve fully closed position learning device according to claim 1, wherein when the determination is made, the energization stop control is executed, and then the fully closed control is executed. 3. Target opening setting means for setting the target opening of the valve;
Feedback control means for feedback-controlling the current to be supplied to the motor so that the detected opening of the valve becomes the set target opening;
The open direction deviation determination means is configured to detect the open direction deviation when the detected opening degree of the valve is equal to or less than a predetermined value near 0 in the state where the target opening is set to 0 during the feedback control. It is determined that the valve has not occurred, and when the detected opening degree of the valve exceeds the predetermined value even after the second predetermined time has elapsed, it is determined that the deviation in the opening direction has occurred. The fully closed position learning device for a valve according to claim 2. The full-closed control means executes the full-closed control when a battery that is a power source of the valve opening degree detecting means is replaced or charged.
The battery closing full-closed position learning means for learning the full-closed position of the valve based on the detected opening degree of the valve detected during the full-closed control is further provided. The valve fully closed position learning device according to any one of the above. The full-closed control means is detected in a state where energization to the motor is stopped along with the stop of the control of the gas flow rate when there is an existing learning value of the full-closed position of the valve. When the degree of deviation between the detected opening of the valve and the existing learning value is equal to or greater than a predetermined threshold value, the full-closed control is executed,
5. A fully-closed position re-learning means for re-learning the fully-closed position of the valve based on the detected opening degree of the valve detected during the fully-closed control. The valve fully closed position learning device according to any one of the above.   After calculating the current value of the fully closed position learning value based on the detected opening of the valve detected during the fully closed control, the fully closed position learning means at the time of battery cancellation is set to the existing learned value. Instead, the current value of the calculated learning value is set as a new learning value, and the fully closed position learning means and the fully closed position relearning means are configured to calculate the existing learning value and the current value of the learning value. 6. The valve fully closed position learning device according to claim 5, wherein both are used to calculate a new learning value.

Images