JP2009002256A - Air-fuel ratio control device of internal combustion engine - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an air-fuel ratio control device of an internal combustion engine free to restrain deterioration of emission and drivability by carrying out F/B re-learning at an optimal time. <P>SOLUTION: Fuel cut is promoted by limiting renewal of a sub F/B learning value (step 112) when sub F/B learning is completed. Thereafter, a sub F/B correction quantity is read (step 114), and re-learning of the sub F/B learning value is executed by releasing the completion of the sub F/B learning (step 118) when the sub F/B correction quantity is more than a prescribed region. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an air-fuel ratio control apparatus for an internal combustion engine.

  An apparatus is known that includes a main air-fuel ratio sensor upstream of an exhaust purification catalyst (hereinafter abbreviated as “catalyst”) and a sub-oxygen sensor downstream of the catalyst. In this apparatus, as air-fuel ratio feedback control, main feedback control based on the output of the main air-fuel ratio sensor (hereinafter referred to as “main F / B control”) and sub-feedback control based on the output of the sub oxygen sensor (hereinafter referred to as “sub F”). / B control ").

Further, immediately after returning from the fuel cut (also referred to as “F / C”) of the internal combustion engine, the oxygen storage amount of the catalyst often reaches a saturated state. In order to restore the purification performance of the catalyst, when returning from the fuel cut, a control (hereinafter referred to as “rich control”) is executed to increase the fuel injection amount and to make the air-fuel ratio richer than the stoichiometric air-fuel ratio. ing.
When such rich control is executed, calculation of feedback correction amount (hereinafter referred to as “F / B correction amount”) by main F / B control or sub F / B control and learning of the F / B correction amount are prohibited. An apparatus has been proposed (see, for example, Patent Document 1). According to this apparatus, erroneous learning of the F / B correction amount during rich control can be prevented.

JP 2005-61356 A JP 2005-105834 A

  According to the apparatus disclosed in Patent Document 1, there is a possibility that the opportunity to perform learning of the F / B correction amount may be reduced. In particular, when the fuel cut is frequently performed after the internal combustion engine is started, there is a situation in which the internal combustion engine is continuously operated without learning the F / B correction amount even once. May be reduced.

  Therefore, when the learning of the F / B correction amount is incomplete, there is a method of restricting the fuel cut execution even if the fuel cut execution condition is satisfied and continuing to update the F / B correction amount learning. Conceivable. According to this method, once the learning of the F / B correction amount is completed, the subsequent fuel cut is allowed to be performed, and the update of the learning of the F / B correction amount is restricted, so that the fuel cut is more than necessary. Therefore, fuel consumption can be prevented from deteriorating.

  However, even if such a method is used, if the learning of the F / B correction amount is completed in an erroneous learning state, the update of learning is restricted, and thus the erroneous learning state continues for a long time. As a result, the air-fuel ratio controllability is deteriorated, which may cause deterioration of exhaust emission characteristics and drivability.

  The present invention has been made in order to solve the above-described problems, and by performing F / B relearning at an optimal timing, it is possible to suppress the deterioration of the emission and drivability of the internal combustion engine. An object of the present invention is to provide a fuel ratio control device.

In order to achieve the above object, a first invention is an air-fuel ratio control apparatus for an internal combustion engine,
A catalyst provided in the exhaust passage of the internal combustion engine;
An exhaust gas sensor provided at least one of upstream and downstream of the catalyst;
Feedback means for performing feedback control based on the output of the exhaust gas sensor to set the air-fuel ratio to the target air-fuel ratio;
Learning means for learning a control center of the feedback control;
Determining means for determining completion or incomplete learning by the learning means;
A fuel cut limiting means for limiting the fuel cut of the internal combustion engine and facilitating the update of the learning when the determination means determines that learning has not been completed;
A learning limiting means for restricting the update of the learning and promoting the fuel cut when learning completion is determined by the determining means;
A learning completion determination canceling unit for canceling the learning completion determination when the learning control unit restricts learning update and the correction amount of the feedback control does not fall within a predetermined range. Features.

The second invention is the first invention, wherein
The feedback means includes main feedback means for executing main feedback control for setting the air-fuel ratio upstream of the catalyst as a target air-fuel ratio based on the output of a main exhaust gas sensor provided upstream of the catalyst, and a sub-feedback provided downstream of the catalyst. Sub-feedback means for performing sub-feedback control for setting the air-fuel ratio downstream of the catalyst to the target air-fuel ratio based on the output of the exhaust gas sensor;
The learning completion determination canceling unit cancels the learning completion determination when the correction amount of the main feedback control or the sub feedback control does not fall within a predetermined range.

A third invention is an air-fuel ratio control apparatus for an internal combustion engine,
A catalyst provided in the exhaust passage of the internal combustion engine;
An exhaust gas sensor provided at least one of upstream and downstream of the catalyst;
Feedback means for executing feedback control for setting the air-fuel ratio to the target air-fuel ratio based on the output of the exhaust gas sensor;
Learning means for learning a control center of the feedback control;
Determining means for determining completion or incomplete learning by the learning means;
A fuel cut limiting means for limiting the fuel cut of the internal combustion engine and facilitating the update of the learning when the determination means determines that learning has not been completed;
A learning limiting means for restricting the update of the learning and promoting the fuel cut when learning completion is determined by the determining means;
After the learning completion determination is made, a learning request value change acquisition unit that acquires a change in a learning request value that is a request value for the learning;
A learning completion determination canceling unit for canceling the learning completion determination when the learning request value change is acquired by the learning request value change acquiring unit when learning update is limited by the learning limiting unit; It is characterized by having.

Moreover, 4th invention is set in 3rd invention,
The learning request value change acquisition means acquires the change of the learning request value based on a fuel property value or an air density.

  In the first invention, when the learning completion determination is made, the learning update by the learning means is restricted and the fuel cut is promoted. Thereby, since fuel cut is not restrict | limited more than needed, the deterioration of a fuel consumption can be prevented. Furthermore, when the update of learning is restricted, the learning completion determination is canceled if the correction amount of the feedback control does not fall within the predetermined range. Thereby, it will be in the state by which learning incomplete determination was made, and the update of learning is accelerated | stimulated. That is, since the relearning by the learning means is executed, the driving is not continuously executed in the erroneous learning state. Therefore, according to the first aspect, it is possible to suppress deterioration of emission and drivability by performing relearning of the control center of feedback control at an optimal timing.

  In the second invention, the learning completion determination is canceled when the correction amount of the main feedback control or the sub feedback control does not fall within the predetermined range. Thereby, relearning of the control center of feedback control can be performed at an optimal timing.

  In the third invention, when the learning completion determination is made, the learning update by the learning means is restricted and the fuel cut is promoted. Thereby, since fuel cut is not restrict | limited more than needed, the deterioration of a fuel consumption can be prevented. Further, when learning update is restricted and learning request value change is acquired, the learning completion determination is canceled. Thereby, it will be in the state by which learning incomplete determination was made, and the update of learning is accelerated | stimulated. That is, since relearning by the learning means is executed, the driving is not continuously executed in a learning state that does not correspond to the latest learning request value. Therefore, according to the third aspect, it is possible to suppress deterioration of emission and drivability by performing relearning of the control center of feedback control at an optimal timing.

  In the fourth invention, a change in the learning request value after the learning completion determination is made is acquired based on the fuel property value or the air density. Therefore, the necessity for re-learning by the learning means can be determined based on the change in the fuel property value or the air density.

  Embodiments of the present invention will be described below with reference to the drawings. In addition, the same code | symbol is attached | subjected to the element which is common in each figure, and the overlapping description is abbreviate | omitted.

Embodiment 1 FIG.
[Description of system configuration]
FIG. 1 is a diagram for explaining the configuration of a system according to Embodiment 1 of the present invention. The system shown in FIG. 1 includes an engine as the internal combustion engine 1. The engine 1 has a plurality of cylinders 2. FIG. 1 shows only one cylinder among a plurality of cylinders.

  The engine 1 includes a cylinder block 6 having a piston 4 inside. The piston 4 is connected to the crankshaft 8 via a crank mechanism. A crank angle sensor 10 that detects a rotation angle (crank angle CA) of the crankshaft 8 is provided in the vicinity of the crankshaft 8.

  A cylinder head 12 is assembled to the upper part of the cylinder block 6. A space from the upper surface of the piston 4 to the cylinder head 12 forms a combustion chamber 14. The cylinder head 12 is provided with a spark plug 16 that ignites the air-fuel mixture in the combustion chamber 14.

  The cylinder head 12 includes an intake port 18 that communicates with the combustion chamber 14. An intake valve 20 is provided at a connection portion between the intake port 18 and the combustion chamber 14. The intake valve 20 is connected to a mechanical or hydraulic variable valve mechanism 22 that can change the valve opening characteristics of the intake valve 20. An intake passage 24 is connected to the intake port 18. An intake passage 24 in the vicinity of the intake port 18 is provided with an injector 26 for injecting fuel in the vicinity thereof. A surge tank 28 is provided upstream of the injector 26.

  A throttle valve 30 is provided upstream of the surge tank 28. The throttle valve 30 is an electronically controlled valve that is driven by a throttle motor 32. The throttle valve 30 is driven based on the accelerator opening AA detected by the accelerator opening sensor 36. A throttle opening sensor 34 that detects the throttle opening TA is provided in the vicinity of the throttle valve 30. An air flow meter 40 is provided upstream of the throttle valve 30. The air flow meter 40 is configured to detect an intake air amount Ga.

  The cylinder head 12 includes an exhaust port 42 that communicates with the combustion chamber 14. An exhaust valve 44 is provided at a connection portion between the exhaust port 42 and the combustion chamber 14. The exhaust valve 44 is connected to a mechanical or hydraulic variable valve mechanism 46 that can change the valve opening characteristics of the exhaust valve 44. An exhaust passage 48 is connected to the exhaust port 44. The exhaust passage 48 is provided with a catalyst 50 that purifies the exhaust gas.

  A main air-fuel ratio sensor (hereinafter referred to as “air-fuel ratio sensor”) 52 that emits a linear output with respect to the exhaust air-fuel ratio is provided upstream of the catalyst 50. A sub-oxygen sensor (hereinafter referred to as “oxygen sensor”) 54 that rapidly changes the output depending on whether the exhaust air-fuel ratio is richer or leaner than the stoichiometric air-fuel ratio is provided downstream of the catalyst 50. ing.

  The system shown in FIG. 1 includes an ECU (Electronic Control Unit) 60 as a control device. An ignition plug 16, variable valve mechanisms 22, 46, an injector 26, a throttle motor 32, and the like are connected to the output side of the ECU 60. On the input side of the ECU 60, there are a crank angle sensor 10, a throttle opening sensor 34, an accelerator opening sensor 36, an air flow meter 40, an air-fuel ratio sensor 52, an oxygen sensor 54, and a fuel property sensor 56 for detecting a fuel property value. Is connected. The ECU 60 controls the entire engine by driving each actuator based on the output of each sensor.

Further, the ECU 60 calculates the engine speed NE based on the crank angle CA. Further, the ECU 60 calculates the engine load KL based on the accelerator opening AA, the throttle opening TA, and the like.
Further, for example, when the accelerator opening AA is zero and the engine speed NE is higher than the reference speed, the ECU 60 executes a fuel cut (F / C) that stops fuel injection from the injector 26.

[Features of Embodiment 1]
In the above system, the ECU 60 executes air-fuel ratio feedback control for setting the air-fuel ratio to the target air-fuel ratio.
FIG. 2 is a block diagram for explaining air-fuel ratio feedback control.
The ECU 60 has a main F / B unit 62 and a sub F / B unit 64. Based on the output of the air-fuel ratio sensor 52, the main F / B means 62 performs main F / B control with the air-fuel ratio upstream of the catalyst 50 as the target air-fuel ratio. Further, the main F / B means 62 stores the average value of the main F / B correction amount calculated by the main F / B control (that is, the control center of the main F / B control) as the main F / B learning value. To do.

  Based on the output of the oxygen sensor 54, the sub F / B means 64 performs sub F / B control with the air / fuel ratio downstream of the catalyst 50 as the target air / fuel ratio. Specifically, sub F / B control is performed to set the output of the oxygen sensor 54 to a target voltage corresponding to the target air-fuel ratio. Further, the sub F / B means 64 stores the average value of the sub F / B correction amount calculated by the sub F / B control (that is, the control center of the sub F / B control) as the sub F / B learning value. To do.

  Although the details will be described later, the ECU 60 executes the update process of the sub F / B learning value until the sub F / B learning value converges. Further, when the sub F / B learning value has converged, the ECU 60 stores the completion of the sub F / B learning in the completion history.

  Further, the ECU 60 calculates the basic fuel injection amount based on the engine load KL. This basic fuel injection amount is corrected by multiplying the F / B correction amount. The F / B correction amount is obtained, for example, by adding the main F / B correction amount and the sub F / B correction amount. Fuel is injected from the injector 26 by the corrected fuel injection amount and supplied into the combustion chamber 14 of the engine 1. In this way, feedback control of the air-fuel ratio is performed.

  By the way, in the above-described Patent Document 1, learning of the F / B correction amount is prohibited during the rich control that is executed immediately after the fuel cut is restored. As a result, there may occur a situation in which the operation of the internal combustion engine is continuously executed without completing the F / B correction amount.

  In order to prevent such a situation, when learning of the F / B correction amount is incomplete, priority is given to the completion of learning of the F / B correction amount by limiting the fuel cut and the rich control after the recovery. A method is conceivable. According to this method, after the learning of the F / B correction amount is once completed, the fuel cut and the rich control after the recovery are preferentially executed by restricting the update of the F / B learning value. Thereby, since execution of a fuel cut is not restrict | limited more than needed, the deterioration of a fuel consumption can be prevented.

  However, when the learning of the F / B correction amount is completed in the erroneous learning state, if the update of the F / B learning value is restricted as described above, the driving state with the erroneously learned F / B learning value Will continue for a long time. As a result, the air-fuel ratio controllability is deteriorated, which may cause deterioration of exhaust emission characteristics and drivability.

  Therefore, in the first embodiment, after the sub F / B learning is completed, if the sub F / B correction amount exceeds the predetermined range, it is determined that the sub F / B learning value is due to erroneous learning. Then, the sub F / B learning value is relearned.

[Specific Processing in Embodiment 1]
FIG. 3 is a flowchart showing a routine executed by the ECU 60 in the first embodiment. This routine is started at predetermined intervals.

  According to the routine shown in FIG. 3, first, the stored value of the completion history of sub F / B learning is read (step 100). Next, it is determined whether or not sub F / B learning is incomplete (step 102). If it is determined in step 102 that the sub F / B learning is not completed, that is, if the completion history does not store the completion of the sub F / B learning, the execution of the fuel cut is restricted (step 104). . In this step 104, lock-up of ECT (Electric Controlled Transmission) can be restricted. Thereby, the update of the sub F / B learning is prioritized over the fuel cut.

  Next, sub F / B learning update processing is executed (step 106). In this step 106, the sub F / B learning value is updated. Thereafter, it is determined whether or not the updated sub F / B learning value has converged (step 108). In this step 108, when the difference between the sub F / B learning value updated in step 106 and the sub F / B learning value before update (updated last time) is equal to or less than the reference value, the sub F / B It is determined that the learning value has converged. If it is determined in step 108 that the sub F / B learning value has not converged, this routine is temporarily terminated.

  On the other hand, if it is determined in step 108 that the sub F / B learning value has converged, the completion of sub F / B learning is stored in the completion history (step 110). Thereafter, this routine is temporarily terminated.

  Thereafter, when this routine is started again, it is determined in step 102 that the sub F / B learning has been completed, and therefore the process proceeds to step 112. In step 112, sub F / B learning is limited. Thereby, the fuel cut is prioritized over the update of the sub F / B learning. Note that the calculation (update) of the sub F / B correction amount is continuously performed.

  Next, the current sub F / B correction amount is read (step 114). Thereafter, it is determined whether or not the sub F / B correction amount read in step 114 exceeds a predetermined range (step 116). In step 116, it is determined whether or not the sub F / B learning value is due to erroneous learning based on the sub F / B correction amount. When it is determined in step 116 that the sub F / B correction amount exceeds the predetermined range, it is determined that the sub F / B learning value is due to erroneous learning, and the completion of the sub F / B learning is completed. Release (step 118). In step 118, the completion of sub F / B learning is cleared from the completion history. As a result, when this routine is started thereafter, YES determination is made at step 102, so that the sub F / B learning value is updated. That is, the relearning of the sub F / B learning value is performed.

As described above, in the first embodiment, when the sub F / B learning is not completed, the fuel cut is limited and the sub F / B learning value is updated with priority. Further, when the sub F / B learning is completed, the update of the sub F / B learning value is restricted, and the fuel cut is promoted. Thereby, since fuel cut is not restrict | limited more than needed, the deterioration of a fuel consumption can be prevented.
Further, in the first embodiment, when the update of the sub F / B learning value is restricted, if the sub F / B correction amount does not fall within the predetermined range, the sub F / B learning value is erroneously learned. Therefore, the completion of the sub F / B learning is canceled. Thereby, since the sub-F / B relearning is executed, the sub-F / B learning value is not continuously executed in a state of erroneous learning. Accordingly, since the relearning of the sub F / B learning value can be performed at the optimum timing, it is possible to suppress the deterioration of emission and drivability.

  In the first embodiment, the engine 1 is the “internal combustion engine” in the first invention, the exhaust passage 48 is the “air-fuel ratio sensor” in the first invention, and the air-fuel ratio sensor 52 is in the first invention. The “exhaust gas sensor” and the “main exhaust gas sensor” in the second invention correspond to the “exhaust gas sensor” in the first invention and the “sub exhaust gas sensor” in the second invention, respectively. In the first embodiment, the main F / B means 62 is the “feedback means” in the first invention and the “main feedback means” in the second invention, and the sub F / B means 64 is the first invention. Corresponds to “feedback means” in FIG. 2 and “sub-feedback means” in the second invention. In the first embodiment, the ECU 60 executes the process of step 106 so that the “learning means” in the first invention executes the processes of steps 108 and 110. When the “determination means” executes the process of step 104, the “fuel cut limiting means” in the first invention executes the process of step 112, and the “learning limiting means” in the first invention executes the process of step 112. , 118 are executed, the “learning completion determination canceling means” in the first and second inventions is realized.

Embodiment 2. FIG.
Next, a second embodiment of the present invention will be described with reference to FIG. 4 and FIG.
The system of the second embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 5 described later using the hardware configuration shown in FIG.

[Features of Embodiment 2]
In the first embodiment, when the sub F / B correction amount exceeds the predetermined range, it is determined that the sub F / B learning value is due to erroneous learning, and the sub F / B learning value is re-learned. To do.

By the way, even when the update of the sub F / B learning is restricted, the update of the main F / B learning can be executed. Also, as shown in FIG. 2, there is a complementary relationship between the main F / B correction amount and the sub F / B correction amount. For this reason, when the update of the sub F / B learning value is restricted, the main F / B correction amount and the main F / B learning value change. Therefore, when the sub F / B learning value is due to erroneous learning, the change in the main F / B learning value becomes large.
Also, the main F / B learning value is stored for each driving region, unlike the sub F / B learning value. In the second embodiment, based on the main F / B learning value, it is determined whether or not the sub F / B learning value is due to erroneous learning.

FIG. 4 is a diagram for explaining a change in the main F / B learning value after completion of the sub F / B learning.
In the second embodiment, when the sub F / B learning is completed, the main F / B learning value KG is set corresponding to the operation areas Ga1 to Ga4 determined by the intake air amount Ga, as indicated by a circle in FIG. Remember. When the sub F / B learning is completed, the update of the main F / B learning value is continuously performed although the update of the sub F / B learning value is limited.

  The square mark in FIG. 4 represents the main F / B learning value after completion of the sub F / B learning. As shown in FIG. 4, in the operation region Ga1, the amount of change in the main F / B learning value from when the sub F / B learning is completed exceeds the reference value. In such a case, in the second embodiment, it is determined that the sub F / B learning value is due to erroneous learning, and the relearning of the sub F / B learning value is executed.

[Specific Processing in Second Embodiment]
FIG. 5 is a flowchart showing a routine executed by ECU 60 in the second embodiment.
According to the routine shown in FIG. 5, as in the routine shown in FIG. 3, if it is determined in step 102 that the sub F / B learning is not completed, the fuel cut is limited (step 104). Update processing of F / B learning is executed (step 106).

  Further, similarly to the routine shown in FIG. 3, if it is determined in step 108 that the sub F / B learning value has converged, the completion of sub F / B learning is stored in the completion history (step 110).

  Thereafter, unlike the routine shown in FIG. 3, the main F / B learning value at the completion of the sub F / B learning is stored in accordance with the operation region (step 120). In step 120, as in the example shown in FIG. 4, the main F / B learning value KG is stored in accordance with the operating region defined by the intake air amount Ga. Then, this routine is temporarily terminated.

  Thereafter, when this routine is started again, it is determined in step 102 that the sub F / B learning has been completed, and therefore the process proceeds to step 112. If the sub-F / B learning update process is restricted in step 112, the fuel cut is prioritized.

  Next, the current main F / B learning value is read (step 122). And the main F / B learning value at the time of completion of sub F / B learning of the corresponding driving | operation area | region is read (step 124). In step 124, the main F / B learning value stored in step 120 is read from the corresponding operation region. Thereafter, the change amount of the main F / B learning value from the time when the sub F / B learning is completed is calculated (step 126). In this step 126, the difference between the current main F / B learning value read in step 122 and the main F / B learning value at the completion of F / B learning read in step 124 is calculated.

  Next, it is determined whether or not the change amount of the main F / B learning value calculated in step 126 is equal to or greater than a reference value (step 128). In step 128, it is determined based on the amount of change in the main F / B learning value whether the sub F / B learning value is due to erroneous learning. If it is determined in step 128 that the change amount of the main F / B learning value is smaller than the reference value, it is determined that the sub F / B learning value is not due to erroneous learning. In this case, this routine is finished as it is.

  On the other hand, if it is determined in step 128 that the change amount of the main F / B learning value is greater than or equal to the reference value, that is, if the change amount of the main F / B learning value does not fall within the predetermined range, It is determined that the F / B learning value is due to erroneous learning. In this case, the completion of the sub F / B learning stored in step 110 is canceled (step 118). In this step 118, the sub F / B learning completion history is cleared. As a result, when this routine is started thereafter, YES determination is made at step 102, so that the sub F / B learning value is updated. That is, the sub F / B learning value is relearned.

As described above, in the second embodiment, when the update of the sub F / B learning value is restricted and the change amount of the main F / B learning value is equal to or larger than the reference value, the sub F / B learning value is updated. It is determined that the B learning value is due to erroneous learning, and the completion of sub F / B learning is canceled. Thereby, since the sub-F / B relearning is executed, the sub-F / B learning value is not continuously executed in a state of erroneous learning. Accordingly, since the relearning of the sub F / B learning value can be performed at the optimum timing, it is possible to suppress the deterioration of emission and drivability.
In addition, since the main F / B learning value is stored for each operation region, whether the sub F / B learning value is due to erroneous learning by using the change amount of the main F / B learning value as a reference. It is possible to judge whether or not systematically.

  In the second embodiment, the “learning completion determination canceling means” according to the second aspect of the present invention is realized by the ECU 60 executing the processing of steps 128 and 118.

Embodiment 3 FIG.
Next, a third embodiment of the present invention will be described with reference to FIG.
The system of the third embodiment can be realized by causing the ECU 60 to execute a routine shown in FIG. 6 described later using the hardware configuration shown in FIG.

[Features of Embodiment 3]
In the first and second embodiments, when the sub F / B learning value is due to erroneous learning, the completion of the sub F / B learning is canceled and the sub F / B learning value is re-learned.

  By the way, for example, when the fuel property value changes due to refueling or the like, the required value for sub F / B learning (hereinafter referred to as “learning required value”) changes. In such a case, if the sub-F / B learning value stored before the change of the fuel property value is continuously used, the air-fuel ratio controllability is lowered, and there is a possibility that exhaust emission characteristics and drivability are deteriorated. is there.

  Therefore, in the third embodiment, when the fuel property value changes more than the reference value after the completion of the sub F / B learning, it is determined that the learning request value has changed, and the sub F / B learning value is re-learned. Execute.

[Specific Processing in Embodiment 3]
FIG. 6 is a flowchart showing a routine executed by the ECU 60 in the third embodiment.
According to the routine shown in FIG. 6, as in the routine shown in FIG. 3, when it is determined in step 102 that the sub F / B learning is not completed, the fuel cut is limited (step 104). Update processing of the F / B learning value is executed (step 106).

  Thereafter, similarly to the routine shown in FIG. 3, if it is determined in step 108 that the sub F / B learning value has converged, the completion of the sub F / B learning is stored (step 110).

  Next, unlike the routine shown in FIG. 3, the fuel property value at the time when the sub F / B learning is completed is stored (step 130). Then, this routine is temporarily terminated.

  Thereafter, when this routine is started again, it is determined in step 102 that the sub-F / B learning has been completed, so that the sub-F / B learning update process is limited as in the routine shown in FIG. (Step 112). By the process of step 112, the fuel cut is prioritized over the sub F / B learning.

  Next, the current fuel property value is read (step 132). Then, the fuel property value at the time when the sub F / B learning is completed is read (step 134). In step 134, the fuel property value stored in step 130 is read. Thereafter, the change amount of the fuel property value from the time when the sub F / B learning is completed is calculated (step 136). In step 136, the difference between the current fuel property value read in step 132 and the fuel property value at the completion of sub F / B learning read in step 134 is calculated.

  Next, it is determined whether or not the change amount of the fuel property value calculated in step 135 is greater than or equal to a reference value (step 138). In this step 138, based on the amount of change in the fuel property value, it is determined whether or not the learning request value has changed after completion of the sub F / B learning. If it is determined in step 138 that the change amount of the fuel property value is smaller than the reference value, it is determined that the required learning value has not changed even after the sub F / B learning is completed. In this case, this routine is finished as it is.

  On the other hand, if it is determined in step 138 that the change amount of the fuel property value is greater than or equal to the reference value, it is determined that the required learning value has changed after the completion of the sub F / B learning. In this case, the completion of the sub F / B learning stored in step 110 is canceled (step 118). As a result, when this routine is started thereafter, YES determination is made at step 102, so that the sub F / B learning value is updated. That is, the sub F / B learning value is relearned.

  By the way, in the third embodiment, whether or not the learning request value has changed after completion of the sub F / B learning is determined based on the change amount of the fuel property value, but based on the change amount of other physical quantities. May be determined. For example, it may be determined whether or not the learning request value has changed based on the amount of change in air density that changes depending on the altitude.

FIG. 7 is a flowchart showing a routine executed by the ECU 60 in the modification of the third embodiment.
According to the routine shown in FIG. 7, unlike the routine shown in FIG. 6, the estimated air density when the sub F / B learning is completed is stored (step 140). Here, the estimated air density value is defined in advance in a map based on the relationship between the throttle opening degree TA and the intake air amount Ga at the opening degree TA, and the map is stored in the ECU 60. In step 140, the estimated air density value obtained by referring to the map when the sub F / B learning is completed is stored.
Note that the air density estimation method is not limited to the method of referring to the map, and a method obtained by calculation can be used.

  Thereafter, when this routine is started again, it is determined in step 102 that the sub-F / B learning has been completed, so that the sub-F / B learning update process is limited as in the routine shown in FIG. (Step 112). By the process of step 112, the fuel cut is prioritized over the sub F / B learning.

  Next, the current air density is estimated (step 142). Then, the air density estimation value at the completion of the sub F / B learning stored in step 140 is read (step 144). Thereafter, a change amount of the air density estimated value from the time when the sub F / B learning is completed is calculated (step 146). In step 146, the difference between the air density estimated value estimated in step 142 and the air density estimated value at the completion of sub F / B learning read in step 144 is calculated.

  Next, it is determined whether or not the change amount of the air density estimated value calculated in step 146 is greater than or equal to a reference value (step 148). In this step 148, it is determined whether or not the learning request value has changed after the completion of the sub F / B learning based on the change amount of the air density estimated value. If it is determined in step 148 that the change amount of the air density estimation value is greater than or equal to the reference value, it is determined that the learning request value has changed after the completion of the sub F / B learning. In this case, the completion of the sub F / B learning stored in step 110 is canceled (step 118). As a result, when this routine is started thereafter, YES determination is made at step 102, so that the sub F / B learning value is updated. That is, the sub F / B learning value is relearned.

As described above, in the third embodiment and the modification, when the sub F / B learning is not completed, the fuel cut is limited, and the sub F / B learning value is updated with priority. The Further, when the sub F / B learning is completed, the update of the sub F / B learning value is restricted, and the fuel cut is promoted. Thereby, since fuel cut is not restrict | limited more than needed, the deterioration of a fuel consumption can be prevented.
Further, in the third embodiment and the modification, when the update of the sub F / B learning value is restricted, if the change amount of the fuel property value or the air density estimation value is equal to or larger than the reference value, the learning request It is determined that the value has changed, and the completion of sub F / B learning is canceled. Accordingly, since the sub-F / B relearning is executed, the driving is not continuously executed in the learning state that does not correspond to the latest learning request value. Accordingly, since the relearning of the sub F / B learning value can be performed at the optimum timing, it is possible to suppress the deterioration of emission and drivability.

  In the third embodiment and the modification, the ECU 60 executes the process of step 106, so that the “learning means” in the first invention executes the processes of steps 108 and 110. The “determination means” in the invention executes the process of step 104, and the “fuel cut limiting means” in the first invention executes the process of step 112, and the “learning limiting means” in the first invention executes the process of step 112. By executing the processing of steps 132 to 136, the “learning request value change acquisition means” in the third and fourth inventions performs “learning completion determination” in the third invention by executing the processing of steps 138 and 118. The “releasing means” executes the processing of steps 148 and 118, whereby the “learning completion determination canceling means” in the fourth invention Each has been realized.

It is a figure for demonstrating the structure of the system by Embodiment 1 of this invention. It is a block diagram for demonstrating air fuel ratio feedback control. 5 is a flowchart showing a routine that is executed by the ECU 60 in the first embodiment of the present invention. It is a figure for demonstrating the change of the main F / B learning value after the completion of sub F / B learning. In Embodiment 2 of this invention, it is a flowchart which shows the routine which ECU60 performs. In Embodiment 3 of this invention, it is a flowchart which shows the routine which ECU60 performs. 10 is a flowchart showing a routine executed by ECU 60 in a modification of the third embodiment of the present invention.

Explanation of symbols

1 Engine 48 Exhaust passage 50 Catalyst 52 Air-fuel ratio sensor 54 Oxygen sensor 56 Fuel property sensor 60 ECU
62 Main F / B means 64 Sub F / B means

Claims (4)

A catalyst provided in the exhaust passage of the internal combustion engine;
An exhaust gas sensor provided at least one of upstream and downstream of the catalyst;
Feedback means for performing feedback control based on the output of the exhaust gas sensor to set the air-fuel ratio to the target air-fuel ratio;
Learning means for learning a control center of the feedback control;
Determining means for determining completion or incomplete learning by the learning means;
A fuel cut limiting means for limiting the fuel cut of the internal combustion engine and facilitating the update of the learning when the determination means determines that learning has not been completed;
A learning limiting means for restricting the update of the learning and promoting the fuel cut when learning completion is determined by the determining means;
A learning completion determination canceling unit for canceling the learning completion determination when the learning control unit restricts learning update and the correction amount of the feedback control does not fall within a predetermined range. An air-fuel ratio control apparatus for an internal combustion engine characterized by the above. The air-fuel ratio control apparatus for an internal combustion engine according to claim 1,
The feedback means includes main feedback means for executing main feedback control for setting the air-fuel ratio upstream of the catalyst as a target air-fuel ratio based on the output of a main exhaust gas sensor provided upstream of the catalyst, and a sub-feedback provided downstream of the catalyst. Sub-feedback means for performing sub-feedback control for setting the air-fuel ratio downstream of the catalyst to the target air-fuel ratio based on the output of the exhaust gas sensor;
The learning completion determination canceling unit cancels the learning completion determination when the correction amount of the main feedback control or the sub feedback control does not fall within a predetermined range. A catalyst provided in the exhaust passage of the internal combustion engine;
An exhaust gas sensor provided at least one of upstream and downstream of the catalyst;
Feedback means for executing feedback control for setting the air-fuel ratio to the target air-fuel ratio based on the output of the exhaust gas sensor;
Learning means for learning a control center of the feedback control;
Determining means for determining completion or incomplete learning by the learning means;
A fuel cut limiting means for limiting the fuel cut of the internal combustion engine and facilitating the update of the learning when the determination means determines that learning has not been completed;
A learning limiting means for restricting the update of the learning and promoting the fuel cut when learning completion is determined by the determining means;
After the learning completion determination is made, a learning request value change acquisition unit that acquires a change in a learning request value that is a request value for the learning;
A learning completion determination canceling unit for canceling the learning completion determination when the learning request value change is acquired by the learning request value change acquiring unit when learning update is limited by the learning limiting unit; An air-fuel ratio control apparatus for an internal combustion engine, comprising: The air-fuel ratio control apparatus for an internal combustion engine according to claim 3,
The learning request value change acquisition means acquires the change of the learning request value based on a fuel property value or an air density.

Images