JP2010168905A - Air-fuel ratio learning control device for internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To learn and correct the displacement of injection characteristics of two fuel injection valves of each cylinder, in an internal combustion engine having two fuel injection valves on the intake side of each cylinder. <P>SOLUTION: While an air-fuel ratio F/B control is performed by injecting both of two fuel injection valves 21 of each cylinder for each fuel injection timing of the cylinder, the displacements (total displacements of injection characteristics) of the injection characteristics of all the fuel injection valves 21 of an engine 11 are learned, based on the corrected amount of the air-fuel ratio F/B. Then, a control to inject a fuel of an injection amount requested by the cylinder to be learned by only one fuel injection valve 21 with the injection of the other fuel injection valve 21 stopped, is performed on only either cylinder (cylinder to be learned) while the air-fuel ratio F/B control is performed while alternately switching between the cylinder to be learned and the injection-stopped fuel injection valve 21. Based on the variation amount of the correction amounts of the air-fuel ratio F/B obtained before and after the stop of the injection of each fuel injection valve 21, the displacement of the injection characteristics of each fuel injection valve 21 of the cylinder to be learned (displacement of individual injection characteristics) is learned individually. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an air-fuel ratio learning control device for an internal combustion engine in which a plurality of fuel injection valves are arranged on the intake side of each cylinder of the internal combustion engine.

  In the conventional general air-fuel ratio control of an internal combustion engine, an exhaust gas sensor for detecting the air-fuel ratio or rich / lean of exhaust gas is arranged upstream of the catalyst in the exhaust passage, and the output of this exhaust gas sensor is used. By executing air-fuel ratio feedback control that sets the air-fuel ratio feedback correction amount and corrects the fuel injection amount of the fuel injection valve of each cylinder with the air-fuel ratio feedback correction amount, the exhaust gas air-fuel ratio is made higher in exhaust gas purification efficiency. The catalyst is controlled within the purification window (near the theoretical air-fuel ratio).

  Further, as described in Japanese Patent Application Laid-Open No. 60-90944, the average injection characteristics of the fuel injection valves of all cylinders based on the air-fuel ratio feedback correction amount during execution of the air-fuel ratio feedback control. A learning correction amount for correcting a deviation (an average air-fuel ratio deviation caused by a deviation of the actual injection amount with respect to the required injection amount) is learned and stored in a memory, and after completion of the fuel injection of each cylinder There is a type in which the fuel injection amount of the valve is corrected by the air-fuel ratio feedback correction amount and the learning correction amount to improve the air-fuel ratio control accuracy.

  Further, as described in Patent Document 2 (Japanese Patent Application Laid-Open No. 2006-299945), atomization of fuel spray in the cylinder of the internal combustion engine and port wet reduction (reduction of fuel adhesion to the inner wall surface of the intake port) For the purpose, for example, there are fuel injection valves arranged in two intake ports of each cylinder of an internal combustion engine, and fuel is injected into each cylinder by two fuel injection valves.

JP-A-60-90944 JP 2006-299945 A

  By the way, due to individual differences (manufacturing variation) of fuel injectors, changes with time, and the like, deviations in the injection characteristics of the fuel injectors occur. Moreover, the deviations in the injection characteristics differ for each fuel injector. Thus, even in a system in which two fuel injection valves are arranged in each cylinder described in Patent Document 2, it is conceivable to apply the air-fuel ratio learning control technique of Patent Document 1 described above. Since the air-fuel ratio learning control technology of this type learns the average injection characteristic deviation (average air-fuel ratio deviation) of the fuel injection valves of all cylinders as a learning correction amount, two fuel injection valves for each cylinder. It is not possible to individually correct each of the injection characteristic deviations. For this reason, problems such as variations in air-fuel ratio between cylinders due to variations in the injection characteristic deviations of the fuel injection valves and variations in torque between the cylinders, and reduction in exhaust gas purification efficiency cannot be solved.

  Therefore, the problem to be solved by the present invention is to correct the learning characteristic correction of the injection characteristics of the plurality of fuel injection valves in each cylinder individually or in a group in an internal combustion engine in which a plurality of fuel injection valves are arranged on the intake side of each cylinder. An object of the present invention is to provide an air-fuel ratio learning control device for an internal combustion engine.

  In order to solve the above-mentioned problem, the invention according to claim 1 is arranged such that a plurality of fuel injection valves are arranged on the intake side of each cylinder of an internal combustion engine having a plurality of cylinders, and an air-fuel ratio of exhaust gas or An exhaust gas sensor for detecting rich / lean is arranged, an air-fuel ratio feedback correction amount is set based on the output of the exhaust gas sensor, and the fuel injection amounts of the plurality of fuel injection valves of each cylinder are set to the air-fuel ratio feedback correction amount. In an air-fuel ratio learning control device for an internal combustion engine having air-fuel ratio control means for executing air-fuel ratio feedback control to be corrected, the fuel injection valves of each cylinder are all injected at every fuel injection timing of each cylinder. When the air-fuel ratio feedback control is being executed, the fuel injection valve injection characteristic deviation of the whole internal combustion engine (hereinafter referred to as the fuel-air-fuel ratio feedback correction amount) The entire injection characteristic deviation learning means for learning "total injection characteristic deviation" and any one cylinder (hereinafter referred to as "learning target cylinder") when the air-fuel ratio feedback control is being executed. Control is performed to stop the injection of one fuel injection valve and inject the fuel for the required injection amount of the learning target cylinder only with the other fuel injection valve, and the learning target cylinder and the fuel injection valve for stopping the injection in order An injection characteristic deviation of each fuel injection valve of the learning target cylinder (hereinafter referred to as “individual injection characteristic deviation”) is performed based on a change amount of the air-fuel ratio feedback correction amount before and after the injection stop of each fuel injection valve. Individual injection characteristic deviation learning means for individually learning, wherein the air-fuel ratio control means uses the learning value for the overall injection characteristic deviation and the learned value for the individual injection characteristic deviation for each cylinder. It is characterized in that to correct the fuel injection amount of the number of fuel injection valves individually.

  In this configuration, during execution of the air-fuel ratio feedback control, all of the fuel injection valves of each cylinder are injected to learn the fuel injection valve injection characteristic deviation (overall injection characteristic deviation) of the entire internal combustion engine, and also the air-fuel ratio feedback control During execution, only one of the cylinders (learning target cylinder) stops injection of any one of the fuel injection valves, and only the other fuel injection valves inject fuel for the required injection amount of the learning target cylinder. In order to learn individually the injection characteristic deviation (individual injection characteristic deviation) of the fuel injection valve, the injection characteristic deviation (individual injection characteristic) of each fuel injection valve of the learning target cylinder while injecting fuel for the required injection amount into the learning target cylinder Deviation) can be individually learned, and the individual injection characteristic deviation can be accurately learned while suppressing variations in air-fuel ratio and torque between the learning target cylinder and other cylinders. Further, in order to individually correct the fuel injection amounts of the plurality of fuel injection valves of each cylinder using the learning value of the overall injection characteristic deviation and the learning value of the individual injection characteristic deviation, the injection characteristic deviation of each fuel injection valve is individually determined. The correction can be performed with high accuracy, and the air-fuel ratio variation and the torque variation between the cylinders due to the variation in the injection characteristic deviation between the fuel injection valves can be effectively reduced.

  In this case, as described in claim 2, after learning of the total injection characteristic deviation is completed by the total injection characteristic deviation learning means, the fuel injection amount of each fuel injection valve is individually corrected in a state where the learning value of the total injection characteristic deviation is corrected. It is preferable to learn the injection characteristic deviation. In this way, since the individual injection characteristic deviation can be learned in a state where the overall injection characteristic deviation is corrected, the learning accuracy of the individual injection characteristic deviation can be further improved.

  However, according to the present invention, the individual injection characteristic deviation is learned without correcting the fuel injection amount of each fuel injection valve with the learning value of the total injection characteristic deviation, and the entire injection is performed after the learning of the total injection characteristic deviation and the individual injection characteristic deviation is completed. The fuel injection amount of each fuel injection valve may be individually corrected using the learned value of characteristic deviation and the learned value of individual injection characteristic deviation.

  In addition to the overall injection characteristic deviation learning means for learning the overall injection characteristic deviation as in claim 3, any one of the plurality of fuel injection valves of each cylinder when the air-fuel ratio feedback control is executed. Control that stops the injection of the fuel injection valve group consisting of one fuel injection valve and injects fuel for the required injection amount of each cylinder only by the other fuel injection valve groups, and in order of the fuel injection valve groups that stop the injection A group that learns an injection characteristic deviation for each fuel injection group (hereinafter referred to as “group-specific injection characteristic deviation”) based on the amount of change in the air-fuel ratio feedback correction amount before and after the injection stop of the fuel injection valve group. A separate injection characteristic deviation learning means is provided, and fuel injection amounts of a plurality of fuel injection valves of each cylinder are injected using the learning value of the overall injection characteristic deviation and the learned value of the individual group injection characteristic deviation. Another to may be corrected group.

  In this configuration, during execution of the air-fuel ratio feedback control, all of the fuel injection valves of each cylinder are injected to learn the fuel injection valve injection characteristic deviation (overall injection characteristic deviation) of the entire internal combustion engine, and also the air-fuel ratio feedback control During the execution, the injection of any one fuel injection valve group is stopped, and only the other fuel injection valve groups inject fuel for the required injection amount of each cylinder. The injection characteristic deviation can be learned while injecting fuel for the required injection amount of each cylinder, and the injection characteristic by group while suppressing air-fuel ratio variation and torque variation between cylinders. The deviation can be learned with high accuracy. Then, in order to correct the fuel injection amount of the plurality of fuel injection valves of each cylinder by the fuel injection valve group using the learning value of the overall injection characteristic deviation and the learning value of the group-specific injection characteristic deviation, the injection of each fuel injection group The characteristic deviation can be corrected with high accuracy, and the air-fuel ratio fluctuation and the torque fluctuation between the cylinders due to the variation in the injection characteristic deviation between the fuel injection valve groups can be reduced.

  In this case, as described in claim 4, after learning of the total injection characteristic deviation is completed by the total injection characteristic deviation learning means, the fuel injection amount of each fuel injection valve is corrected with the learning value of the total injection characteristic deviation. It is preferable to learn the different injection characteristic deviation. In this way, since the group-specific injection characteristic deviation can be learned in a state where the overall injection characteristic deviation is corrected, the learning accuracy of the group-specific injection characteristic deviation can be further improved.

  However, the present invention learns the group-specific injection characteristic deviation in a state where the fuel injection amount of each fuel injection valve is not corrected with the learning value of the total injection characteristic deviation, and after completing the learning of the total injection characteristic deviation and the group-specific injection characteristic deviation The fuel injection amount of each fuel injection valve may be corrected for each fuel injection valve group using the learning value of the overall injection characteristic deviation and the learning value of the group-specific injection characteristic deviation.

  Further, as in claim 5, the overall injection characteristic deviation, the group-specific injection characteristic deviation, and the individual injection characteristic deviation are learned, respectively, and the learning value of the total injection characteristic deviation, the learning value of the group-specific injection characteristic deviation, and the individual injection characteristic deviation The fuel injection amount of the plurality of fuel injection valves of each cylinder may be corrected using the learned value. In this way, it is possible to reduce the variation in air-fuel ratio between cylinders and the variation in torque caused by the variation in the injection characteristic deviation between the fuel injection valves and the variation in the injection characteristic deviation between the fuel injection valve groups.

  In this case, as described in claim 6, after learning of the total injection characteristic deviation is completed by the whole injection characteristic deviation learning unit, and after learning of the group-specific injection characteristic deviation is completed by the group-specific injection characteristic deviation learning unit, The individual injection characteristic deviation may be learned in a state where the fuel injection amount of the fuel injection valve is corrected by the learning value of the overall injection characteristic deviation and the learning value of the group-specific injection characteristic deviation. In this way, since the individual injection characteristic deviation can be learned in a state where both the overall injection characteristic deviation and the group-specific injection characteristic deviation are corrected, the learning accuracy of the individual injection characteristic deviation can be further improved.

  Further, as in claim 7, when the learning value is stored in a rewritable storage means that retains stored data even when the internal combustion engine is stopped, and the learning value is stored in the storage means, the storage It is preferable to correct the fuel injection amounts of the plurality of fuel injection valves of each cylinder using the learned value read from the means. In this way, when the learning value is stored in the storage means, it is possible to correct the fuel injection amounts of the plurality of fuel injection valves of each cylinder using the learning value from the start of the internal combustion engine. Improved startability and reduced emissions at start-up.

  Further, as in claim 8, there may be provided an abnormality determining means for determining that the fuel injection valve is abnormal and warning the driver when the learned value is out of a predetermined allowable range. In this way, when an abnormality occurs in the fuel injection valve, it becomes possible to specify an abnormal fuel injection valve, or to specify a fuel injection valve group to which the abnormal fuel injection valve belongs. It is possible to promptly notify the driver that a valve abnormality has occurred and prompt repair and inspection.

FIG. 1 is a schematic configuration diagram of an entire engine control system used in common in the first to third embodiments of the present invention. FIG. 2 is a diagram illustrating a configuration example of a four-cylinder engine that is commonly used in the first to third embodiments of the present invention. FIG. 3 is a flowchart showing the flow of processing of the injection characteristic deviation learning correction main routine of the first embodiment. FIG. 4 is a flowchart showing the flow of processing of the entire injection characteristic deviation learning routine of the first embodiment. FIG. 5 is a flowchart showing the flow of processing of the individual injection characteristic deviation learning routine of the first embodiment. FIG. 6 is a flowchart showing the flow of processing of the learning control routine of the first embodiment. FIG. 7 is a diagram for explaining the variation in the injection characteristics of the fuel injection valve. FIG. 8 is a flowchart showing the flow of processing of the injection characteristic deviation learning correction main routine of the second embodiment. FIG. 9 is a flowchart showing the flow of processing of the group-specific injection characteristic deviation learning routine of the second embodiment. FIG. 8 is a flowchart showing the flow of processing of the injection characteristic deviation learning correction main routine of the third embodiment.

  Hereinafter, three embodiments 1 to 3 embodying the mode for carrying out the present invention will be described.

A first embodiment of the present invention will be described with reference to FIGS.
First, a schematic configuration of the entire engine control system will be described with reference to FIG.
An air cleaner 13 is provided at the most upstream portion of the intake pipe 12 of the engine 11 that is an internal combustion engine, and an air flow meter 14 that detects the intake air amount is provided downstream of the air cleaner 13. A throttle valve 16 whose opening is adjusted by a motor 15 and a throttle opening sensor 17 for detecting the opening (throttle opening) of the throttle valve 16 are provided on the downstream side of the air flow meter 14.

  Further, a surge tank 18 is provided on the downstream side of the throttle valve 16, and an intake pipe pressure sensor 19 for detecting the intake pipe pressure is provided in the surge tank 18. An intake manifold 20 that introduces air into each cylinder is connected between the surge tank 18 and the intake port 31 of each cylinder of the engine 11. On the other hand, a spark plug 22 is attached to the cylinder head of the engine 11 for each cylinder, and the air-fuel mixture in the cylinder is ignited by the spark discharge of each spark plug 22.

  As shown in FIG. 2, each cylinder of the engine 11 is provided with two intake ports 31 and two exhaust ports 32, and in the vicinity of the two intake ports 31 of each cylinder, toward the intake port 31, respectively. A fuel injection valve 21 for injecting fuel is disposed, and two fuel injection valves 21 are disposed per cylinder. Each intake port 31 is opened and closed by an intake valve 33, and each exhaust port 32 is opened and closed by an exhaust valve 34.

  On the other hand, as shown in FIG. 1, the exhaust pipe 23 (exhaust passage) of the engine 11 is provided with an exhaust gas sensor 24 (air-fuel ratio sensor, oxygen sensor, etc.) for detecting the air-fuel ratio or rich / lean of the exhaust gas. A catalyst 25 such as a three-way catalyst for purifying exhaust gas is provided downstream of the exhaust gas sensor 24.

  A cooling water temperature sensor 26 that detects the cooling water temperature and a knock sensor 29 that detects knocking vibration are attached to the cylinder block of the engine 11. A crank angle sensor 28 that outputs a pulse signal every time the crankshaft 27 rotates by a predetermined crank angle is attached to the outer peripheral side of the crankshaft 27. Based on the output pulse signal of the crank angle sensor 28, the crank angle and The engine speed is detected.

  Outputs of these various sensors are input to an engine control circuit (hereinafter referred to as “ECU”) 30. The ECU 30 is mainly composed of a microcomputer, and executes various engine control programs stored in a built-in ROM (storage medium), so that the two fuel injection valves 21 of each cylinder according to the engine operating state. The fuel injection amount of the fuel injection valve 21 and the ignition timing of the spark plug 22 are controlled.

  Further, when a predetermined air-fuel ratio F / B control execution condition is satisfied, the ECU 30 makes the air-fuel ratio of the exhaust gas coincide with the target air-fuel ratio (for example, the theoretical air-fuel ratio) based on the output of the exhaust gas sensor 24. Air-fuel ratio control means that calculates an air-fuel ratio F / B correction amount and executes air-fuel ratio F / B control for correcting the fuel injection amount of the two fuel injection valves 21 of each cylinder using the air-fuel ratio F / B correction amount. Function as. Here, “F / B” means “feedback” (hereinafter the same).

  By the way, as shown in FIG. 7, the injection characteristic deviation (deviation of the actual injection amount with respect to the required injection amount) of the fuel injection valve 21 occurs due to the individual difference (manufacturing variation) of the fuel injection valve 21 and the change with time, and the like. The deviation of the injection characteristic differs for each fuel injection valve 21.

  In consideration of this point, the ECU 30 executes each routine for correcting the injection characteristic deviation learning shown in FIGS. 3 to 6 described later, and when predetermined learning execution conditions are satisfied, the fuel injection of the entire engine 11 is performed. The overall injection characteristic deviation learning for learning the injection characteristic deviation of the valve 21 (hereinafter referred to as “total injection characteristic deviation”) and the injection characteristic deviation for each fuel injection valve 21 (hereinafter referred to as “individual injection characteristic deviation”) are individually learned. The individual injection characteristic deviation learning is executed, and the fuel injection amounts of the two fuel injection valves 21 of each cylinder are individually corrected using the learning value of the overall injection characteristic deviation and the learned value of the individual injection characteristic deviation. Yes.

  Here, in the overall injection characteristic deviation learning, the air-fuel ratio F / B is controlled when the air-fuel ratio F / B control is executed by injecting all the two fuel injection valves 21 of each cylinder at every fuel injection timing of each cylinder. Based on the B correction amount, the injection characteristic deviation (total injection characteristic deviation) of the fuel injection valve 21 of the entire engine 11 is learned.

  On the other hand, in the individual injection characteristic deviation learning, during the execution of the air-fuel ratio F / B control, the injection of the fuel injection valve 21 on one side (for example, the left side) is stopped in only one of the cylinders (hereinafter referred to as “learning target cylinder”). Then, the control for injecting the fuel for the required injection amount of the learning target cylinder only by the other (for example, the right side) fuel injection valve 21 is performed by sequentially switching the learning target cylinder and the fuel injection valve 21 for stopping the injection. Thus, the injection characteristic deviation (individual injection characteristic deviation) of each fuel injection valve 21 of the learning target cylinder is individually learned based on the change amount of the air-fuel ratio F / B correction amount before and after the injection stop of each fuel injection valve 21.

  Hereinafter, the processing contents of each routine for correcting the injection characteristic deviation learning of FIGS. 3 to 6 executed by the ECU 30 will be described. In the following description, the four-cylinder engine shown in FIG. 2 will be described as an example, and the numbers of the respective cylinders are denoted by # 1, # 2, # 3, and # 4 as necessary, and the respective cylinders # 1 to # 4 are described. These two fuel injection valves 21 are represented by A and B.

[Injection characteristic deviation learning correction main routine]
The injection characteristic deviation learning correction main routine of FIG. 3 is repeatedly executed periodically during engine operation. When this routine is started, first, in step 101, whether or not a predetermined learning execution condition is satisfied is determined, for example, by [1] being in steady operation (for example, in idle operation) and [2] Judgment is made based on whether a condition such as the execution of air-fuel ratio F / B control is satisfied. As a result, if it is determined that the learning execution condition is not satisfied, it is determined that the operation condition is not suitable for the learning of the injection characteristic deviation, and the processing relating to the learning of the injection characteristic deviation from step 102 onward is performed. This routine ends.

  On the other hand, if it is determined in step 101 that the learning execution condition is satisfied, it is determined that the operation condition is suitable for learning the injection characteristic deviation, and the processing related to learning of the injection characteristic deviation in step 102 and subsequent steps is performed. Run as follows: First, in step 102, an overall injection characteristic deviation learning routine of FIG. 4 described later is executed, and an injection characteristic deviation (total injection characteristic deviation) of the fuel injection valve 21 of the entire engine 11 is based on the air-fuel ratio F / B correction amount AFFB. ).

  Thereafter, the process proceeds to step 103, where it is determined whether or not the overall injection characteristic deviation learning has been completed (whether or not the final overall injection characteristic deviation learning value Gf.all.fin = 1), and the overall injection characteristic deviation learning is completed. Until then, the learning of the overall injection characteristic deviation is continued.

  Thereafter, when the entire injection characteristic deviation learning is completed, the routine proceeds to step 104 where the entire injection characteristic deviation learned value Gf.all is updated and stored in the backup RAM 38 which is a rewritable storage means for retaining the stored data even when the engine is stopped. The entire injection characteristic deviation learning value Gf.all is applied to the air-fuel ratio control. Thus, during the execution of the air-fuel ratio control, the fuel injection amounts of the fuel injection valves 21 of all the cylinders are uniformly corrected with the total injection characteristic deviation learning value Gf.all.

  Thereafter, the routine proceeds to step 105, where an individual injection characteristic deviation learning routine shown in FIG. 5 to be described later is executed to switch the learning target cylinder and the fuel injection valve 21 for stopping the injection in order, and stop the injection of each fuel injection valve 21. Based on the change amount of the front / rear air-fuel ratio F / B correction amount, the injection characteristic deviation (individual injection characteristic deviation) of the fuel injection valve 21 that stopped the injection of the learning target cylinder is individually learned.

  Thereafter, the process proceeds to step 106, where it is determined whether or not the individual injection characteristic deviation learning is completed (whether or not the fuel injection valve identification number NUM = 9 [final number]), and until the individual injection characteristic deviation learning is completed. Continue learning the individual injection characteristic deviation.

  Thereafter, when the individual injection characteristic deviation learning is completed, the routine proceeds to step 107 where the individual injection characteristic deviation learned values Gf. # 1A to Gf. # 4A and Gf. # 1B to Gf. # 4B are updated and stored in the backup RAM 38. These individual injection characteristic deviation learning values Gf. # 1A to Gf. # 4A and Gf. # 1B to Gf. # 4B are applied to the air-fuel ratio control. Thereby, during the execution of the air-fuel ratio control, the individual fuel injection characteristic deviation learning values Gf. # 1A to Gf. # 4A, Gf. Corresponding to the fuel injection amounts of the fuel injection valves A and B of the cylinders # 1 to # 4, respectively. Corrected by # 1B to Gf. # 4B. Here, Gf. # 1A to Gf. # 4A are individual injection characteristic deviation learning values for individually correcting the fuel injection amount of one fuel injection valve A of each of the cylinders # 1 to # 4. . # 1B to Gf. # 4B are individual injection characteristic deviation learning values for individually correcting the fuel injection amount of the other fuel injection valve B of each of the cylinders # 1 to # 4.

[Overall injection characteristic deviation learning routine]
The whole injection characteristic deviation learning routine of FIG. 4 is a subroutine executed in step 102 of the injection characteristic deviation learning correction main routine of FIG. 3 and plays a role as whole injection characteristic deviation learning means in the claims. When this routine is started, first, in step 201, the entire injection characteristic deviation learned value Gf.all read from the backup RAM 38 is set to the air-fuel ratio learned value Gf. As a result, during the execution of the overall injection characteristic deviation learning, the fuel injection amounts of the fuel injection valves 21 of all the cylinders are uniformly corrected with the air-fuel ratio learning value Gf (total injection characteristic deviation learned value Gf.all). If the total injection characteristic deviation learned value Gf.all is not stored in the backup RAM 38, a preset initial value is set as the air-fuel ratio learned value Gf.

  Thereafter, the routine proceeds to step 202, where a learning control routine of FIG. 6 described later is executed to update the air-fuel ratio learned value Gf, and then the routine proceeds to step 203 where the current air-fuel ratio learned value Gf is changed to the entire injection characteristic deviation learned value Gf. Set to .all and update and store in the backup RAM 38.

  Thereafter, the process proceeds to step 204, in which it is determined whether or not learning is completed (whether the air-fuel ratio F / B correction amount AFFB is substantially 0), and the processes in steps 202 to 204 are repeated until learning is completed. . Thereafter, when it is determined in step 204 that the learning is completed, the process proceeds to step 205 where the final total injection characteristic deviation learned value Gf.all.fin is set to “1”, and this routine is terminated.

  When learning (updating and storing) the total injection characteristic deviation learning value Gf.all, it is determined whether or not the total injection characteristic deviation learning value Gf.all is within a predetermined allowable range. When the characteristic deviation learning value Gf.all is out of the allowable range, it is determined that one of the fuel injection valves 21 is abnormal, and the abnormality information is stored in the backup RAM 38, and a warning lamp is lit or blinked. A warning may be displayed on the display section of the instrument panel on the driver's seat to warn the driver. Further, guard processing is performed so that the total injection characteristic deviation learning value Gf.all falls within a predetermined allowable range (the upper and lower limit values of the total injection characteristic deviation learning value Gf.all are limited to within a predetermined value). May be.

[Individual injection characteristic deviation learning routine]
The individual injection characteristic deviation learning routine of FIG. 5 is a subroutine executed in step 105 of the injection characteristic deviation learning correction main routine of FIG. 3, and plays a role as individual injection characteristic deviation learning means in the claims. When this routine is started, first, at step 301, the following (1) to (8) are performed based on the fuel injection valve identification number NUM (= 1 to 9) counted up at the last step 306 of this routine. And the fuel injection valve 21 for stopping the injection is selected.

  (1) When the fuel injection valve identification number NUM = 1, the cylinder # 1 becomes a learning target cylinder, the injection of one fuel injection valve A of the cylinder # 1 is stopped, and the other fuel injection of the cylinder # 1 By increasing the injection amount of the valve B by a factor of 2, the required amount of fuel is injected into the cylinder # 1.

  (2) When the fuel injection valve identification number NUM = 2, the cylinder # 1 becomes the learning target cylinder, the injection of the other fuel injection valve B of the cylinder # 1 is stopped, and the one fuel injection of the cylinder # 1 By increasing the injection amount of the valve A by a factor of 2, the required amount of fuel is injected into the cylinder # 1.

  (3) When the fuel injection valve identification number NUM = 3, the cylinder # 2 becomes the learning target cylinder, the injection of one fuel injection valve A of the cylinder # 2 is stopped, and the other fuel injection of the cylinder # 2 By increasing the injection amount of the valve B by a factor of 2, the required amount of fuel is injected into the cylinder # 2.

  (4) When the fuel injection valve identification number NUM = 4, the cylinder # 2 becomes the learning target cylinder, the injection of the other fuel injection valve B of the cylinder # 2 is stopped, and the one fuel injection of the cylinder # 2 is stopped By increasing the injection amount of the valve A by a factor of 2, the required amount of fuel is injected into the cylinder # 2.

  (5) When the fuel injection valve identification number NUM = 5, the cylinder # 3 becomes the learning target cylinder, the injection of one fuel injection valve A of the cylinder # 3 is stopped, and the other fuel injection of the cylinder # 3 By increasing the injection amount of the valve B by a factor of 2, the required amount of fuel is injected into the cylinder # 3.

  (6) When the fuel injection valve identification number NUM = 6, the cylinder # 3 becomes the learning target cylinder, the injection of the other fuel injection valve B of the cylinder # 3 is stopped, and the one fuel injection of the cylinder # 3 By increasing the injection amount of the valve A by a factor of 2, the required amount of fuel is injected into the cylinder # 3.

  (7) When the fuel injection valve identification number NUM = 7, cylinder # 4 becomes the learning target cylinder, the injection of one fuel injection valve A of the cylinder # 4 is stopped, and the other fuel injection of the cylinder # 4 By increasing the injection amount of the valve B by a factor of 2, the required amount of fuel is injected into the cylinder # 4.

  (8) When the fuel injection valve identification number NUM = 8, the cylinder # 4 becomes the learning target cylinder, the injection of the other fuel injection valve B of the cylinder # 4 is stopped, and the one fuel injection of the cylinder # 4 By increasing the injection amount of the valve A by a factor of 2, the required amount of fuel is injected into the cylinder # 4.

  Thereafter, the routine proceeds to step 302, where it is determined which of the following applies based on the current fuel injection valve identification number NUM, and the individual injection characteristic deviation learning value corresponding to the current fuel injection valve identification number NUM is empty. Set to the fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 1, the individual injection characteristic deviation learning value Gf. # 1A of one fuel injection valve A of the cylinder # 1 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 2, the individual injection characteristic deviation learning value Gf. # 1B of the other fuel injection valve B of the cylinder # 1 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 3, the individual injection characteristic deviation learning value Gf. # 2A of one fuel injection valve A of the cylinder # 2 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 4, the individual fuel injection characteristic deviation learning value Gf. # 2B of the other fuel injection valve B of the cylinder # 2 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 5, the individual injection characteristic deviation learning value Gf. # 3A of one fuel injection valve A of the cylinder # 3 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 6, the individual injection characteristic deviation learning value Gf. # 3B of the other fuel injection valve B of the cylinder # 3 is set to the air-fuel ratio learning value Gf.

  When the fuel injection valve identification number NUM = 7, the individual injection characteristic deviation learning value Gf. # 4A of one fuel injection valve A of the cylinder # 4 is set to the air-fuel ratio learning value Gf.

  In the case of the fuel injection valve identification number NUM = 8, the individual injection characteristic deviation learning value Gf. # 4B of the other fuel injection valve B of the cylinder # 4 is set to the air-fuel ratio learning value Gf.

  Thereafter, the routine proceeds to step 303, where a learning control routine shown in FIG. 6 described later is executed to update the air-fuel ratio learned value Gf, and then the routine proceeds to step 304 where any of the following is performed based on the current fuel injection valve identification number NUM. It is discriminated whether it is true, and the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value corresponding to the fuel injection valve identification number NUM.

  When the fuel injection valve identification number NUM = 1, the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value Gf. # 1A of one fuel injection valve A of the cylinder # 1.

  When the fuel injection valve identification number NUM = 2, the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value Gf. # 1B of the other fuel injection valve B of the cylinder # 1.

  When the fuel injector identification number NUM = 3, the current air-fuel ratio learned value Gf is set to the individual injection characteristic deviation learned value Gf. # 2A of one fuel injector A of the cylinder # 2.

  When the fuel injector identification number NUM = 4, the current air-fuel ratio learned value Gf is set to the individual injection characteristic deviation learned value Gf. # 2B of the other fuel injector B of the cylinder # 2.

  When the fuel injection valve identification number NUM = 5, the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value Gf. # 3A of one fuel injection valve A of the cylinder # 3.

  When the fuel injector identification number NUM = 6, the current air-fuel ratio learned value Gf is set to the individual injection characteristic deviation learned value Gf. # 3B of the other fuel injector B of the cylinder # 3.

  When the fuel injection valve identification number NUM = 7, the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value Gf. # 4A of one fuel injection valve A of the cylinder # 4.

  When the fuel injection valve identification number NUM = 8, the current air-fuel ratio learning value Gf is set to the individual injection characteristic deviation learning value Gf. # 4B of the other fuel injection valve B of the cylinder # 4.

  Thereafter, the process proceeds to step 305, where it is determined whether learning is completed (whether the air-fuel ratio F / B correction amount AFFB is substantially 0), and the processing of steps 303 to 305 is repeated until learning is completed. . Thereafter, when it is determined in step 305 that learning is completed, the routine proceeds to step 306, where the fuel injection valve identification number NUM is incremented by 1, and this routine is terminated.

  When learning the individual injection characteristic deviation learning values Gf. # 1A to Gf. # 4A and Gf. # 1B to Gf. # 4B, the individual injection characteristic deviation learning values are within a predetermined allowable range. If any individual injection characteristic deviation learning value is out of the allowable range, it is determined that the fuel injection valve 21 out of the allowable range is abnormal, and the abnormality information is stored in the backup RAM 38. In addition, the warning lamp may be turned on or blinked, or a warning may be displayed on the display section of the driver's instrument panel to warn the driver. In this case, when any one of the fuel injection valves 21 is abnormal, the abnormal fuel injection valve 21 can be specified. Further, guard processing may be performed so that the individual injection characteristic deviation learned value falls within a predetermined allowable range (the upper and lower limit values of the individual injection characteristic deviation learned value are limited to within a predetermined value).

[Learning control routine]
The learning control routine of FIG. 6 is a subroutine executed in step 202 of the entire injection characteristic deviation learning routine of FIG. 4 and step 303 of the individual injection characteristic deviation learning routine of FIG. When this routine is started, first, at step 401, the current air-fuel ratio learned value Gf is set to the previous air-fuel ratio learned value Gf.o, and at the next step 402, the current air-fuel ratio F / B correction amount AFFB is set. Is set to the previous air-fuel ratio F / B correction amount AFFB.o.

  Thereafter, it is determined whether the air-fuel ratio F / B correction amount AFFB is a positive value, 0, or a negative value by the determination processing in steps 403 and 404, and the air-fuel ratio F / B correction amount AFFB is determined to be 0. If it has been determined (when both the determinations in steps 403 and 404 are “Yes”), the routine proceeds to step 405, where the air-fuel ratio learning value update amount DLGF is set to zero.

  When the air-fuel ratio F / B correction amount AFFB is determined to be a negative value (when “No” is determined in step 403), the process proceeds to step 407, where the air-fuel ratio learning value update amount DLGF is set in advance. Set to a predetermined value (negative value).

  On the other hand, if it is determined that the air-fuel ratio F / B correction amount AFFB is a positive value (“Yes” in step 403 and “No” in step 404), the process proceeds to step 406 and the air-fuel ratio learning value is reached. The update amount DLGF is set to a predetermined value (plus value) set in advance.

  After the air-fuel ratio learning value update amount DLGF is set as described above, the routine proceeds to step 408, where the value obtained by adding the air-fuel ratio learning value update amount DLGF to the previous air-fuel ratio learning value Gf.o is set as the new air-fuel ratio learning value Gf. And

  Thereafter, the routine proceeds to step 409, where the value obtained by adding the air-fuel ratio learning value update amount DLGF to the previous air-fuel ratio F / B correction amount AFFB.o is set as a new air-fuel ratio F / B correction amount AFFB, and this routine is ended.

  According to the first embodiment described above, when the air-fuel ratio F / B control is executed by injecting all the two fuel injection valves 21 of each cylinder at every fuel injection timing of each cylinder, the air-fuel ratio F After learning the injection characteristic deviation (total injection characteristic deviation) of the fuel injection valve 21 of the entire engine 11 based on the / B correction amount, any one cylinder (learning target cylinder) during the execution of the air-fuel ratio F / B control Only the other fuel injection valve 21 stops the injection of the fuel injection valve 21 and injects fuel corresponding to the required injection amount of the learning target cylinder, and the fuel injection valve 21 stops the learning target cylinder and the injection. Are switched in order, and the injection characteristic deviation (individual injection characteristic deviation) of each fuel injection valve 21 of the learning target cylinder is based on the change amount of the air-fuel ratio F / B correction amount before and after the injection stop of each fuel injection valve 21. ) To learn individually, The injection characteristic deviation (individual injection characteristic deviation) of each fuel injection valve 21 of the learning target cylinder can be individually learned while injecting fuel for the required injection amount into the learning target cylinder. The individual injection characteristic deviation can be learned with high accuracy while suppressing the air-fuel ratio variation and torque variation between the two. Since the fuel injection amounts of the two fuel injection valves 21 of each cylinder are individually corrected using the learning value of the overall injection characteristic deviation and the learned value of the individual injection characteristic deviation, the injection characteristic deviation of each fuel injection valve 21 is corrected. The correction can be made individually with high accuracy, and the variation in air-fuel ratio between the cylinders and the variation in torque caused by the variation in the injection characteristic deviation between the fuel injection valves 21 can be effectively reduced.

  Moreover, in the first embodiment, after learning of the total injection characteristic deviation is completed, the individual injection characteristic deviation is learned in a state where the fuel injection amount of each fuel injection valve 21 is corrected with the learning value of the total injection characteristic deviation. Therefore, the individual injection characteristic deviation can be learned with high accuracy in a state where the overall injection characteristic deviation is corrected, and the learning accuracy of the individual injection characteristic deviation can be further improved.

  However, in the present invention, the individual injection characteristic deviation is learned in a state where the fuel injection amount of each fuel injection valve 21 is not corrected by the learning value of the total injection characteristic deviation, and the entire injection characteristic deviation and the individual injection characteristic deviation are learned after completing the learning. The fuel injection amount of each fuel injection valve 21 may be individually corrected using the learned value of the injection characteristic deviation and the learned value of the individual injection characteristic deviation.

  Next, Embodiment 2 of the present invention will be described with reference to FIGS. However, description of substantially the same parts as those in the first embodiment will be omitted or simplified, and different parts from the first embodiment will be mainly described.

  In the second embodiment, after learning the total injection characteristic deviation by the same method as in the first embodiment, the group-specific injection characteristic deviation is learned, and the learning value of the total injection characteristic deviation and the learning value of the group-specific injection characteristic deviation are The fuel injection amount of the two fuel injection valves 21 of each cylinder is corrected for each fuel injection group.

Here, in learning of the injection characteristic deviation by group, during the execution of the air-fuel ratio F / B control, the fuel injection valve group including one (for example, the left side) fuel injection valve 21 of the two fuel injection valves 21 of each cylinder. The fuel injection valve is controlled by switching the fuel injection valve group for stopping the injection and controlling the injection of the fuel for the required injection amount of each cylinder only by the other fuel injection valve group (for example, the right side). A fuel injection valve group-specific injection characteristic deviation (group-specific injection characteristic deviation) is learned based on the amount of change in the air-fuel ratio F / B correction amount before and after stopping the injection of the group.
The processing contents of the routines for correcting the injection characteristic deviation learning in FIGS. 8 and 9 executed in the second embodiment will be described below.

[Injection characteristic deviation learning correction main routine]
The injection characteristic deviation learning correction main routine of FIG. 8 is repeatedly executed periodically during engine operation. When this routine is started, first, in steps 501 to 504, air-fuel ratio F / B correction is performed by the same processing as steps 101 to 104 of the injection characteristic deviation learning correction main routine of FIG. 3 described in the first embodiment. The injection characteristic deviation (total injection characteristic deviation) of the fuel injection valve 21 of the entire engine 11 is learned based on the amount AFFB, and when the whole injection characteristic deviation learning is completed, the backup RAM 38 stores the total injection characteristic deviation learned value Gf. All is updated and stored, and this total injection characteristic deviation learning value Gf.all is applied to the air-fuel ratio control.

  Thereafter, the routine proceeds to step 505, where a group-by-group injection characteristic deviation learning routine of FIG. 9 described later is executed to stop the injection of one fuel injection valve group and the required injection of each cylinder only by the other fuel injection valve group. An amount of fuel is injected, and an injection characteristic deviation (group-specific injection characteristic deviation) for each fuel injection valve group is learned based on the change amount of the air-fuel ratio F / B correction amount before and after the injection stop of the fuel injection valve group. .

  Thereafter, the process proceeds to step 506, where it is determined whether or not the group-specific injection characteristic deviation learning is completed (whether or not the group identification number GRP = 3 [final number]), and until the group-specific injection characteristic deviation learning is completed. Continue learning the injection characteristic deviation by group.

  Thereafter, when the group-specific injection characteristic deviation learning is completed, the routine proceeds to step 507, where the group-specific injection characteristic deviation learned values Gf.A and Gf.B are updated and stored in the backup RAM 38, and this group-specific injection characteristic deviation learned value is obtained. Gf.A and Gf.B are applied to air-fuel ratio control. Thus, during execution of the air-fuel ratio control, the fuel injection amounts of the fuel injection valves A and B of the cylinders # 1,..., # 4 are group-specific injection characteristic deviation learning values Gf.A and Gf.B for each fuel injection group. It is corrected by.

[Group-specific injection characteristic deviation learning routine]
The group-specific injection characteristic deviation learning routine of FIG. 9 is a subroutine executed in step 505 of the injection characteristic deviation learning correction main routine of FIG. 8 and serves as a group-specific injection characteristic deviation learning means in the claims. Fulfill. When this routine is started, first, in step 601, it is determined which of the following applies based on the group identification number GRP counted up in the last step 606 of this routine, and the fuel for executing the injection Select the injection valve group.

  When the group identification number GRP = 1, the fuel injection valve group A (one fuel injection valve A of each cylinder # 1 to # 4) is injected only, and the other fuel injection valve group B (each cylinder # 1). Stop injection of the other fuel injection valve B) of # 4.

  In the case of the group identification number GRP = 2, injection is performed only by the other fuel injection valve group B (the other fuel injection valve B of each cylinder # 1 to # 4), and one fuel injection group A (each cylinder # 1). The injection of one fuel injection valve A) of # 4 is stopped.

  Thereafter, the process proceeds to step 602, where it is determined which of the following applies based on the group identification number GRP, and the group-specific injection characteristic deviation learned value corresponding to the current group identification number GRP is set to the air-fuel ratio learned value Gf. set.

When the group identification number GRP = 1, the group-specific injection characteristic deviation learned value Gf.A of one fuel injection valve group A is set to the air-fuel ratio learned value Gf.
When the group identification number GRP = 2, the group-specific injection characteristic deviation learned value Gf.B of the other fuel injection valve group B is set to the air-fuel ratio learned value Gf.

  Thereafter, the process proceeds to step 603, the learning control routine of FIG. 6 described in the first embodiment is executed to update the air-fuel ratio learned value Gf, and then the process proceeds to step 604, where any of the following is performed based on the group identification number GRP. And the current air-fuel ratio learning value Gf is set to the group-specific injection characteristic deviation learning value corresponding to the group identification number GRP.

When the group identification number GRP = 1, the current air-fuel ratio learned value Gf is set to the group-specific injection characteristic deviation learned value Gf.A of one fuel injection valve group A.
When the group identification number GRP = 2, the current air-fuel ratio learning value Gf is set to the group-specific injection characteristic deviation learning value Gf.B of the other fuel injection valve group B.

  Thereafter, the process proceeds to step 605, where it is determined whether learning has been completed (whether the air-fuel ratio F / B correction amount AFFB is substantially zero), and the processing of steps 603-605 is repeated until learning is completed. . Thereafter, when it is determined in step 605 that learning is completed, the process proceeds to step 606, where the group identification number GRP is incremented by 1, and this routine is terminated.

  When learning the injection characteristic deviation learning values Gf.A and Gf.B for each group, it is determined whether or not the learning characteristic deviation values for each group are within a predetermined allowable range. If the individual injection characteristic deviation learning value is out of the allowable range, it is determined that one of the fuel injection valves 21 in the fuel injection valve group out of the allowable range is abnormal, and the abnormality information is stored in the backup RAM 38. The information may be stored and a warning lamp may be turned on or blinked, or a warning may be displayed on the display unit of the instrument panel of the driver's seat to warn the driver. In this case, when an abnormality occurs in any of the fuel injection valves 21, the fuel injection valve group to which the abnormal fuel injection valve 21 belongs can be specified. Further, guard processing may be performed so that the group-specific injection characteristic deviation learning value falls within a predetermined allowable range (the upper and lower limit values of the group-specific injection characteristic deviation learning value are limited within a predetermined value).

  In the second embodiment described above, during the execution of the air-fuel ratio F / B control, the two fuel injection valves 21 of each cylinder are all injected to perform the injection characteristic deviation (total injection characteristic deviation) of the engine 11 as a whole. ) Is learned, during the execution of the air-fuel ratio F / B control, the fuel injection is stopped while the fuel injection valve group stops the injection and the fuel injection valve only injects fuel for the required injection amount of each cylinder. In order to learn the injection characteristic deviation for each valve group (injection characteristic deviation for each group), it is possible to learn the injection characteristic deviation for each group while injecting fuel corresponding to the required injection amount of each cylinder. In addition, it is possible to accurately learn the injection characteristic deviation by group while suppressing torque variation. Then, in order to correct the fuel injection amount of the two fuel injection valves 21 of each cylinder by the fuel injection valve group using the learning value of the overall injection characteristic deviation and the learning value of the group-specific injection characteristic deviation, It is possible to correct the injection characteristic deviation with high accuracy, and to reduce the air-fuel ratio fluctuation and the torque fluctuation between the cylinders due to the variation in the injection characteristic deviation between the fuel injection valve groups.

  Further, in the second embodiment, after learning of the total injection characteristic deviation is completed, the group-specific injection characteristic deviation is learned in a state where the fuel injection amount of each fuel injection valve 21 is corrected with the learning value of the total injection characteristic deviation. Therefore, the group-specific injection characteristic deviation can be learned in a state where the overall injection characteristic deviation is corrected, and the learning accuracy of the group-specific injection characteristic deviation can be further improved.

  However, the present invention learns the group-specific injection characteristic deviation without correcting the fuel injection amount of each fuel injection valve 21 with the learning value of the total injection characteristic deviation, and completes the learning of the total injection characteristic deviation and the group-specific injection characteristic deviation. Later, the fuel injection amount of each fuel injection valve 21 may be corrected for each fuel injection valve group using the learned value for the overall injection characteristic deviation and the learned value for the group-specific injection characteristic deviation.

  In the third embodiment of the present invention, the overall injection characteristic deviation, the group-specific injection characteristic deviation, and the individual injection characteristic deviation described in the first and second embodiments are respectively learned, and the learning value of the total injection characteristic deviation and the group-specific injection characteristic deviation are learned. The fuel injection amount of the plurality of fuel injection valves 21 of each cylinder is corrected using the learned value of this and the learned value of the individual injection characteristic deviation.

  In the third embodiment, the injection characteristic deviation learning correction main routine of FIG. 10 is periodically and repeatedly executed during engine operation. When this routine is started, first, in steps 701 to 704, the air-fuel ratio F / B correction is performed by the same processing as in steps 101 to 104 of the injection characteristic deviation learning correction main routine of FIG. 3 described in the first embodiment. The injection characteristic deviation (total injection characteristic deviation) of the fuel injection valve 21 of the entire engine 11 is learned based on the amount AFFB, and when the whole injection characteristic deviation learning is completed, the backup RAM 38 stores the total injection characteristic deviation learned value Gf. All is updated and stored, and this total injection characteristic deviation learning value Gf.all is applied to the air-fuel ratio control.

  Thereafter, in steps 705 to 707, the injection of one fuel injection valve group is stopped by the same processing as in steps 505 to 507 of the injection characteristic deviation learning correction main routine of FIG. The fuel for the required injection amount of each cylinder is injected only by the fuel injection valve group, and the injection for each fuel injection group is performed based on the change amount of the air-fuel ratio F / B correction amount before and after the injection stop of the fuel injection group. When the characteristic deviation (group-specific injection characteristic deviation) is learned and the group-specific injection characteristic deviation learning is completed, the group-specific injection characteristic deviation learned values Gf.A and Gf.B are updated and stored in the backup RAM 38, and the group-specific injection characteristic deviation is learned. The injection characteristic deviation learning values Gf.A and Gf.B are applied to the air-fuel ratio control.

  Thereafter, in steps 708 to 710, the learning target cylinder and the fuel injection valve 21 that stops the injection are processed by the same processing as steps 105 to 107 of the injection characteristic deviation learning correction main routine of FIG. 3 described in the first embodiment. The respective injection characteristics deviation (individual injection characteristic deviation) of each fuel injection valve 21 of the learning target cylinder is individually switched based on the change amount of the air-fuel ratio F / B correction amount before and after the injection stop of each fuel injection valve 21. When the individual injection characteristic deviation learning is completed, the individual injection characteristic deviation learned values Gf. # 1A to Gf. # 4A and Gf. # 1B to Gf. # 4B are updated and stored in the backup RAM 38, and the individual injection is performed. The characteristic deviation learning values Gf. # 1A to Gf. # 4A and Gf. # 1B to Gf. # 4B are applied to the air-fuel ratio control.

  According to the third embodiment described above, the overall injection characteristic deviation, the group-specific injection characteristic deviation, and the individual injection characteristic deviation are learned, respectively, and the learning value of the overall injection characteristic deviation, the learning value of the group-specific injection characteristic deviation, and the individual injection Since the fuel injection amount of the plurality of fuel injection valves 21 of each cylinder is corrected using the learned value of the characteristic deviation, the variation in the injection characteristic deviation between the fuel injection valves 21 and the injection characteristic deviation between the fuel injection valve groups. It is possible to reduce air-fuel ratio variations and torque variations between cylinders due to variations in the engine.

  Moreover, in the third embodiment, after learning of the total injection characteristic deviation is completed, the group-specific injection characteristic deviation is learned in a state where the fuel injection amount of each fuel injection valve 21 is corrected with the learning value of the total injection characteristic deviation. Thereafter, the individual injection characteristic deviation is learned in a state where the fuel injection amount of each fuel injection valve 21 is corrected by the learning value of the whole injection characteristic deviation and the learning value of the group-specific injection characteristic deviation. The individual injection characteristic deviation can be learned in a state where both the group-specific injection characteristic deviations are corrected, and the learning accuracy of the individual injection characteristic deviation can be further improved.

However, in the present invention, the order of learning of the overall injection characteristic deviation, learning of the individual-group injection characteristic deviation, and learning of the individual injection characteristic deviation may be appropriately changed.
In addition, the present invention is not limited to the four-cylinder engine shown in FIG. 2, but can be applied to an engine having three or less cylinders or five or more cylinders, or an engine having three or more fuel injection valves in each cylinder. May be.

  DESCRIPTION OF SYMBOLS 11 ... Engine (internal combustion engine), 12 ... Intake pipe, 16 ... Throttle valve, 21 ... Fuel injection valve, 22 ... Spark plug, 23 ... Exhaust pipe (exhaust passage), 24 ... Exhaust gas sensor, 30 ... ECU (Air-fuel ratio control) Means, overall injection characteristic deviation learning means, individual injection characteristic deviation learning means, group-specific injection characteristic deviation learning means, abnormality determination means), 31 ... intake port, 32 ... exhaust port, 38 ... backup RAM (storage means).

Claims (8)

A plurality of fuel injection valves are arranged on the intake side of each cylinder of an internal combustion engine having a plurality of cylinders, and an exhaust gas sensor for detecting the air-fuel ratio or rich / lean of the exhaust gas is arranged in the exhaust passage, and the exhaust gas sensor Air-fuel ratio control means for performing air-fuel ratio feedback control for setting the air-fuel ratio feedback correction amount based on the output and correcting the fuel injection amounts of the plurality of fuel injection valves of each cylinder with the air-fuel ratio feedback correction amount is provided. In an air-fuel ratio learning control device for an internal combustion engine,
The entire internal combustion engine is controlled based on the air-fuel ratio feedback correction amount when the air-fuel ratio feedback control is executed by injecting all of the plurality of fuel injection valves of each cylinder at every fuel injection timing of each cylinder. An overall injection characteristic deviation learning means for learning a fuel injection valve injection characteristic deviation (hereinafter referred to as “total injection characteristic deviation”);
When the air-fuel ratio feedback control is being executed, only one of the cylinders (hereinafter referred to as “learning target cylinder”) stops the injection of any one of the fuel injection valves, and only the other fuel injection valves perform the learning. The control for injecting fuel for the required injection amount of the target cylinder is performed by sequentially switching the learning target cylinder and the fuel injection valve for stopping the injection, and the air-fuel ratio feedback correction before and after the injection stop of each fuel injection valve Individual injection characteristic deviation learning means for individually learning the injection characteristic deviation of each fuel injection valve of the learning target cylinder (hereinafter referred to as “individual injection characteristic deviation”) based on the amount of change in the amount;
The air-fuel ratio control unit individually corrects the fuel injection amounts of the plurality of fuel injection valves of the cylinders using the learning value of the overall injection characteristic deviation and the learned value of the individual injection characteristic deviation. An air-fuel ratio learning control device for an internal combustion engine.   The individual injection characteristic deviation learning means corrects the fuel injection amount of each fuel injection valve with the learning value of the total injection characteristic deviation after the learning of the whole injection characteristic deviation is completed by the whole injection characteristic deviation learning means. The air-fuel ratio learning control device for an internal combustion engine according to claim 1, wherein the individual injection characteristic deviation is learned in a state. A plurality of fuel injection valves are arranged on the intake side of each cylinder of an internal combustion engine having a plurality of cylinders, and an exhaust gas sensor for detecting the air-fuel ratio or rich / lean of the exhaust gas is arranged in the exhaust passage, and the exhaust gas sensor Air-fuel ratio control means for performing air-fuel ratio feedback control for setting the air-fuel ratio feedback correction amount based on the output and correcting the fuel injection amounts of the plurality of fuel injection valves of each cylinder with the air-fuel ratio feedback correction amount is provided. In an air-fuel ratio learning control device for an internal combustion engine,
The entire internal combustion engine is controlled based on the air-fuel ratio feedback correction amount when the air-fuel ratio feedback control is executed by injecting all of the plurality of fuel injection valves of each cylinder at every fuel injection timing of each cylinder. An overall injection characteristic deviation learning means for learning a fuel injection valve injection characteristic deviation (hereinafter referred to as “total injection characteristic deviation”);
When the air-fuel ratio feedback control is being executed, the injection of the fuel injection valve group consisting of any one of the plurality of fuel injection valves of each cylinder is stopped, and only the other fuel injection valve groups The amount of change in the air-fuel ratio feedback correction amount before and after the injection stop of the fuel injection valve group is controlled by sequentially switching the fuel injection valve group that stops the injection to control the injection of the fuel for the required injection amount of each cylinder. And an injection characteristic deviation learning means for each group for learning an injection characteristic deviation for each fuel injection valve group (hereinafter referred to as “group-specific injection characteristic deviation”)
The air-fuel ratio control means corrects the fuel injection amounts of the plurality of fuel injection valves of each cylinder for each fuel injection valve group using the learning value for the overall injection characteristic deviation and the learning value for the group-specific injection characteristic deviation. An air-fuel ratio learning control apparatus for an internal combustion engine, characterized in that:   The group-specific injection characteristic deviation learning unit corrects the fuel injection amount of each fuel injection valve with the learning value of the total injection characteristic deviation after the total injection characteristic deviation learning unit completes the learning of the total injection characteristic deviation. 4. The air-fuel ratio learning control apparatus for an internal combustion engine according to claim 3, wherein the deviation of the injection characteristics by group is learned in a state of being performed. When the air-fuel ratio feedback control is being executed, only one of the cylinders (hereinafter referred to as “learning target cylinder”) stops the injection of any one of the fuel injection valves, and only the other fuel injection valves perform the learning. The control for injecting fuel for the required injection amount of the target cylinder is performed by sequentially switching the learning target cylinder and the fuel injection valve for stopping the injection, and the air-fuel ratio feedback correction before and after the injection stop of each fuel injection valve An individual injection characteristic deviation learning means for individually learning an injection characteristic deviation of each fuel injection valve of the learning target cylinder (hereinafter referred to as “individual injection characteristic deviation”) based on a change amount of the amount;
The air-fuel ratio control means uses the learned value of the overall injection characteristic deviation, the learned value of the group-specific injection characteristic deviation, and the learned value of the individual injection characteristic deviation, and the fuel injection amounts of the plurality of fuel injection valves of the cylinders The air-fuel ratio learning control apparatus for an internal combustion engine according to claim 3 or 4, wherein   The individual injection characteristic deviation learning means has completed learning of the total injection characteristic deviation by the total injection characteristic deviation learning means, and has completed learning of the group-specific injection characteristic deviation by the group-specific injection characteristic deviation learning means. The individual injection characteristic deviation is learned after the fuel injection amount of each fuel injection valve is corrected with the learning value of the overall injection characteristic deviation and the learning value of the group-specific injection characteristic deviation. 6. An air-fuel ratio learning control device for an internal combustion engine according to claim 5. The learning value is stored in a rewritable storage means that retains stored data even when the internal combustion engine is stopped,
When the learning value is stored in the storage unit, the air-fuel ratio control unit corrects the fuel injection amounts of the plurality of fuel injection valves of the cylinders using the learning value read from the storage unit. The air-fuel ratio learning control apparatus for an internal combustion engine according to any one of claims 1 to 6, wherein   8. An abnormality determining unit that determines that the fuel injection valve is abnormal and warns a driver when the learned value is out of a predetermined allowable range. An air-fuel ratio learning control apparatus for an internal combustion engine.

Images