JP2021085362A - Exhaust emission control system - Google Patents

Abstract

To provide an exhaust emission control system for an internal combustion engine, which can avoid erroneous learning of a correction value of an output when an optimum purification point fluctuates even though the output of an air-fuel ratio sensor is correct.SOLUTION: A control device performs air-fuel ratio feedback control based on a deviation between a rear air-fuel ratio and a target air-fuel ratio, and learning processing for learning a correction value of the rear air-fuel ratio. In the learning processing, a learning value is calculated based on the deviation. When the rear air-fuel ratio has a value on a lean side of the target air-fuel ratio, the correction value is updated using the learning value. When the rear air-fuel ratio has a value on a rich side of the target air-fuel ratio, it is determined whether or not a time series value of the deviation fluctuates to the rich side over time. When the time series value is determined not to fluctuate to the rich side, the correction value is updated using the learning value. When the time series value is determined to fluctuate to the rich side, a current value of the rear air-fuel ratio is set as a new target air-fuel ratio in the air-fuel ratio feedback control.SELECTED DRAWING: Figure 6

Description

The present invention relates to a system that purifies exhaust gas from an internal combustion engine.

WO 2012/157111 discloses an air-fuel ratio sensor compensator provided downstream of the catalyst. This compensator performs active control. In active control, the air-fuel ratio of the gas flowing into the catalyst is controlled to oscillate across the stoichiometric air-fuel ratio. This correction device also calculates a correction value based on the deviation between the output of the air-fuel ratio sensor and the reference output during a predetermined period during the execution of active control. This correction value is used to correct the output of the air-fuel ratio sensor in the air-fuel ratio control executed separately from the active control. Therefore, according to this correction device, it is possible to eliminate the variation in the air-fuel ratio sensor and improve the accuracy of the air-fuel ratio control.

International Publication No. 2012/157111

However, in the above-mentioned correction device, it is difficult to deal with the case where the optimum purification point fluctuates due to the deterioration of the catalyst even though the output of the air-fuel ratio sensor is correct. That is, when the optimum purification point fluctuates, the air-fuel ratio sensor correctly outputs the air-fuel ratio corresponding to the optimum purification point during the execution of active control. However, since a deviation occurs between this output and the reference output, the correction value is calculated. Then, during the execution of the air-fuel ratio control, the output of the air-fuel ratio sensor is corrected by this correction value, and there is a possibility that the exhaust gas purification efficiency by the catalyst is lowered. Therefore, improvements are desired to avoid such erroneous learning.

One object of the present invention is to prevent the correction value of the output from being erroneously learned when the output of the air-fuel ratio sensor is correct but the optimum purification point fluctuates.

The present invention is an exhaust gas purification system for an internal combustion engine that solves the above problems, and has the following features.
The exhaust gas purification system includes a catalyst, an air-fuel ratio sensor, and a control device.
The catalyst purifies the exhaust gas from the internal combustion engine.
The air-fuel ratio sensor detects the rear air-fuel ratio, which is the air-fuel ratio of the exhaust gas that has passed through the catalyst.
The control device performs air-fuel ratio feedback control based on the deviation between the rear air-fuel ratio and the target air-fuel ratio, and learning processing for learning the correction value of the rear air-fuel ratio.
The control device is used in the learning process.
The learning value is calculated based on the deviation,
When the rear air-fuel ratio is a value on the lean side of the target air-fuel ratio, the correction value is updated using the learning value.
When the rear air-fuel ratio is a value on the rich side of the target air-fuel ratio, it is determined whether or not the time series value of the deviation including the current value of the deviation fluctuates to the rich side over time.
When it is determined that the time-series value does not fluctuate toward the rich side over time, the learning value is used to update the correction value.
When it is determined that the time-series value fluctuates toward the rich side over time, the current value of the rear air-fuel ratio is set as a new target air-fuel ratio in the air-fuel ratio feedback control.

According to the present invention, when the rear air-fuel ratio is a value on the lean side of the target air-fuel ratio, the correction value of the rear air-fuel ratio is updated using the learned value. Even when the rear air-fuel ratio is a value on the rich side of the target air-fuel ratio and the time series value of the deviation of these air-fuel ratios does not fluctuate to the rich side over time, the learning value is used to make the rear air-fuel ratio. The fuel ratio correction value is updated. On the other hand, if the rear air-fuel ratio is a value on the rich side of the target air-fuel ratio and the time series value fluctuates to the rich side over time, the current value of the rear air-fuel ratio is the air-fuel ratio feedback control. Set to a new target air-fuel ratio in.

If the rear air-fuel ratio is richer than the target air-fuel ratio, the optimum purification point may fluctuate even though the output of the air-fuel ratio sensor is correct. In this regard, according to the present invention, when it is determined that the time series value fluctuates toward the rich side over time, the current value of the rear air-fuel ratio is set to the new target air-fuel ratio without updating the correction value. Is set to. Therefore, according to the present invention, it is possible to avoid erroneous learning of the correction value of the output when the optimum purification point fluctuates even though the output of the air-fuel ratio sensor is correct. Become. In addition, by setting the value this time to the new target air-fuel ratio, it is possible to improve the accuracy of the air-fuel ratio feedback control and improve the exhaust gas purification efficiency by the catalyst.

It is a figure explaining the structural example of the exhaust gas purification system which concerns on embodiment. It is a figure explaining an example of the time to execute the learning process which concerns on embodiment. It is a figure explaining the rear air-fuel ratio when a learning process is executed. It is a figure explaining the rear air-fuel ratio when a learning process is executed. It is a figure explaining an example of the method for determining whether or not the optimum purification point fluctuated to the rich side from the target air-fuel ratio. It is a flowchart which shows the flow when a control device executes a learning process. It is a figure explaining the effect by execution of a learning process.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, when the number, quantity, quantity, range, etc. of each element is referred to in the embodiment shown below, the reference is made unless otherwise specified or clearly specified by the number in principle. The present invention is not limited to the number of the above. In addition, the structures, steps, and the like described in the embodiments shown below are not necessarily essential to the present invention, unless otherwise specified or clearly specified in principle.

1. 1. Configuration of Exhaust Gas Purification System FIG. 1 is a diagram illustrating a configuration example of an exhaust gas purification system for an internal combustion engine according to an embodiment of the present invention. The exhaust gas purification system 10 shown in FIG. 1 is mounted on a vehicle. The exhaust gas purification system 10 includes an internal combustion engine 1 as a power source for the vehicle. Examples of the internal combustion engine 1 include a diesel engine and a gasoline engine.

The internal combustion engine 1 includes an exhaust pipe 2. Exhaust gas from the internal combustion engine 1 flows through the exhaust pipe 2. A three-way catalyst 3 is provided in the middle of the exhaust pipe 2. The three-way catalyst 3 purifies harmful components (for example, HC, CO and NOx) in the exhaust gas when the air-fuel ratio of the exhaust gas flowing into the three-way catalyst 3 is in a narrow range near the stoichiometric air-fuel ratio AFth (for example, 14.60). To do.

An air-fuel ratio sensor 4 is provided upstream of the three-way catalyst 3. The air-fuel ratio sensor 4 outputs a signal proportional to the air-fuel ratio of the exhaust gas flowing into the three-way catalyst 3 (hereinafter, also referred to as “front air-fuel ratio AFfr”). On the other hand, an air-fuel ratio sensor 5 is provided downstream of the three-way catalyst 3. The air-fuel ratio sensor 5 outputs a signal proportional to the air-fuel ratio of the exhaust gas that has passed through the three-way catalyst 3 (hereinafter, also referred to as “rear air-fuel ratio AFrr”). _{These air-fuel ratio sensors are composed of linear detection type O 2} sensors that can continuously detect the air-fuel ratio of exhaust gas over a relatively wide range.

The exhaust gas purification system 10 also includes an ECU (Electronic Control Unit) 6 as a control device. The ECU 6 is a microcomputer provided with a processor, a memory, and an input / output interface. In addition to the air-fuel ratio sensors 4 and 5, various sensors necessary for various controls of the vehicle and the internal combustion engine 1 are connected to the input side of the ECU 6. On the other hand, various actuators such as an injection device (not shown) that injects fuel into the internal combustion engine 1 are connected to the output side of the ECU 6.

The ECU 6 uses the outputs of the air-fuel ratio sensors 4 and 5 to execute the air-fuel ratio feedback control. The air-fuel ratio feedback control includes a main feedback control based on the output of the air-fuel ratio sensor 4 and a sub-feedback control based on the output of the air-fuel ratio sensor 5. In the main feedback control, the main feedback value is calculated based on the deviation Dfr between the front air-fuel ratio AFfr and the theoretical air-fuel ratio AFth. In the sub-feedback control, the sub-feedback value is calculated based on the deviation Drr between the rear air-fuel ratio AFrr and the target air-fuel ratio AFtrg (for example, 14.60) corresponding to the optimum purification point of the three-way catalyst 3. The main and sub-feedback values are reflected in the fuel injection calculation.

2. Learning process 2-1. The premise ECU 6 also performs a learning process for learning the correction value CAF of the rear air-fuel ratio AFrr. The correction value CAF is a value or a coefficient for correcting the rear air-fuel ratio AFrr (AFrr ← AFrr + CAF, AFrr ← AFrr × CAF). The learning process is executed when a predetermined learning condition is satisfied. This learning condition includes preconditions and timing conditions. Prerequisites include that the three-way catalyst 3 is in an active state, that the air-fuel ratio sensors 4 and 5 are in a normal state, and that the operating state of the internal combustion engine 1 is in a state in which air-fuel ratio feedback control can be executed. I'm out.

The timing condition is a condition relating to the timing when the learning process is executed. An example of the time when the learning process is executed is the time when the rear air-fuel ratio AFrr is switched from the lean side value to the rich side value after returning from the fuel cut control. As the time to execute the learning process, the time when the rear air-fuel ratio AFrr is largely switched from the lean side value to the rich side value during the execution of the active control is also exemplified. The timing conditions are set corresponding to these timings. The active control is a control that transiently switches the target air-fuel ratio AFtrg between the value on the rich side and the value on the lean side.

FIG. 2 is a diagram illustrating an example of a time when the learning process is executed. In the example shown in FIG. 2, the learning process is executed after returning from the fuel cut control (F / C control). In the learning process, the deviation Dave between the average value AVE of the output while the output of the air-fuel ratio sensor 5 (that is, the rear air-fuel ratio AFrr) is stable and the target air-fuel ratio AFtrg is calculated as the learning value LAF. When the learning value LAF is calculated, the correction value CAF is updated using this learning value LAF. The average value AVE may be calculated using the output after being processed by the low-pass filter. Instead of the average value AVE, the mode value of the output while the output of the air-fuel ratio sensor 5 is stable may be used.

2-2. Problems of Learning Process FIGS. 3 and 4 are diagrams for explaining the rear air-fuel ratio AFrr when the learning process is executed. In the example shown in FIG. 3, the optimum purification point of the three-way catalyst 3 coincides with the target air-fuel ratio AFtrg. As shown in FIG. 3 (i), the true air-fuel ratio AFrl of the exhaust gas that has passed through the three-way catalyst 3 does not match the optimum purification point. Therefore, it is possible that the purification capacity of the three-way catalyst 3 has not been fully utilized. However, the rear air-fuel ratio AFrr is in agreement with the target air-fuel ratio AFtrg. Therefore, even if the sub-feedback control is executed in such a situation, the sub-feedback value is not calculated, and the purification of harmful components may be insufficient.

However, as shown in FIG. 3 (ii), during the execution of the learning process, the rear air-fuel ratio AFrr when the true air-fuel ratio AFrr coincides with the optimum purification point is obtained. This rear air-fuel ratio AFrr does not match the target air-fuel ratio AFtrg. Therefore, when the learning process is executed, the deviation Dave is calculated as the learning value LAF (LAF = AFrr-AFtrg).

Then, as shown in FIG. 3 (iii), when the correction value CAF is updated using the learning value LAF, the rear air-fuel ratio AFrr is corrected to the rich side by the amount of the correction value CAF. Since the corrected rear air-fuel ratio AFrr (= AFrr + CAF or AFrr × CAF) matches the target air-fuel ratio AFtrg, there is no change in the fuel injection amount based on the sub-feedback value. However, if the sub-feedback control based on the corrected rear air-fuel ratio AFrr is executed, the true air-fuel ratio AFrr can be brought closer to the optimum purification point. Therefore, it is possible to purify harmful components by fully utilizing the purifying ability of the three-way catalyst 3.

Similar to the example shown in FIG. 3, in the example shown in FIG. 4, the true air-fuel ratio AFrl does not match the optimum purification point. Therefore, it is possible that the purification capacity of the three-way catalyst 3 has not been fully utilized. However, as shown in FIG. 4 (i), the rear air-fuel ratio AFrr is in agreement with the target air-fuel ratio AFtrg. Therefore, the problem described in FIG. 3 (i) can also occur in FIG. 4 (i).

However, unlike the example shown in FIG. 3, in the example shown in FIG. 4, the cause of the above problem is that the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg due to the deterioration of the three-way catalyst 3. That is, the rear air-fuel ratio AFrr is output correctly and is not the cause of the above problem. Therefore, when the learning value LAF (LAF = AFrr-AFtrg) is calculated and the correction value CAF is updated as shown in FIG. 4 (ii), a new problem arises.

That is, as shown in FIG. 4 (iii), the rear air-fuel ratio AFrr coincides with the target air-fuel ratio AFtrg. Therefore, even when the sub-feedback control based on the corrected rear air-fuel ratio AFrr is executed, the problem that may occur in FIG. 4 (i) is not solved at all.

2-3. Improvement points of the learning process Therefore, in the learning process according to the embodiment, it is determined whether or not the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg. The reason why the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg is that the adsorption property and purification ability of harmful components are lowered due to the deterioration of the three-way catalyst 3. If so, the more the deterioration progresses, the greater the fluctuation toward the rich side. This tendency is judged based on the time series value TS of the deviation Dave.

FIG. 5 is a diagram illustrating an example of a method for determining whether or not the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg based on the time series value TS. FIG. 5 shows the number of times of learning processing k (k = 1 to 13) and the average value AVE at each time. The number k of the learning process indicates that the smaller the number, the older the number.

In the explanation of FIG. 5, it is assumed that the target air-fuel ratio Trg is set to the theoretical air-fuel ratio AFth (14.60). First, it is assumed that the number of times k of the learning process this time is k = 4 (that is, it is assumed that the learning process after k = 5 has not been executed).

In this case (first case), each of the deviation Dave between the average value AVE of k = 1 to 4 and the target air-fuel ratio AFtrg is the time series value TS. The average value AVE coincides with the target air-fuel ratio AFtrg at k = 1, 2, and 4, and shows a value on the lean side at k = 3. Therefore, in the first case, it is determined that the optimum purification point does not fluctuate to the rich side of the target air-fuel ratio AFtrg.

Next, it is assumed that the number of times k of the learning process this time is k = 8 (that is, the learning process after k = 9 is not executed). In this case, each of the average value AVE of k = 1 to 8 and the deviation Dave from the target air-fuel ratio AFtrg is the time series value TS. You may choose k = 5-8 by reducing the number going back from this value.

Here, the case where k = 5 to 8 is selected (second case) will be described. In the second case, the average value AVE shows all the values on the rich side. However, the average value AVE increases from k = 5 to k = 6. That is, the average value AVE moves to the lean side with time from k = 5 to k = 6. Therefore, in the second case, it is determined that the optimum purification point does not fluctuate to the rich side of the target air-fuel ratio AFtrg.

Next, it is assumed that the number of times k of the learning process this time is k = 13. In this case, each of the deviation Dave between the average value AVE of k = 1 to 13 and the target air-fuel ratio AFtrg is the time series value TS. The number going back from the value this time may be reduced, and k = 8 to 13 may be selected, or k = 10 to 13 may be selected.

Here, the case where k = 10 to 13 is selected (third case) will be described. In the third case, the average value AVE shows all the values on the rich side. In the third case, the average value AVE also fluctuates toward the rich side with time from k = 10 to k = 13. Therefore, in the third case, since the deviation Dave becomes smaller with time, it is determined that the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg. Since the average value AVE did not increase at least from k = 11 to k = 12, it is considered that the deviation Dave decreased with time from k = 10 to k = 13.

In the learning process according to the embodiment, when it is determined that the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg, the current average value AVE is set to the target air-fuel ratio AFtrg. That is, the current average value AVE is set as a new target air-fuel ratio AFtrg in the sub-feedback control. When it is determined that the optimum purification point does not fluctuate to the rich side of the target air-fuel ratio AFtrg, the learning process described in item 2.1 is executed and the correction value CAF is updated.

2-4. Specific Processing FIG. 6 is a flowchart showing a flow of processing executed by the ECU 6. The processing routine shown in FIG. 6 is repeatedly executed in a predetermined control cycle.

As shown in FIG. 6, the ECU 6 first determines whether the preconditions and timing conditions are satisfied (steps S10 and S12). Examples of these conditions have already been described. If at least one of these conditions is not met, the ECU 6 terminates the processing routine. When both of these conditions are satisfied, the ECU 6 starts executing the learning process.

Subsequently, the ECU 6 calculates the average value AVE (step S14), and determines whether or not the average value AVE matches the target air-fuel ratio AFtrg (step S16). If the determination result in step S16 is affirmative, the ECU 6 ends the processing routine.

If the determination result in step S16 is negative, the ECU 6 determines whether or not the average value AVE is smaller than the target air-fuel ratio AFtrg (step S18). That is, the ECU 6 determines whether or not the average value AVE is a value on the lean side of the target air-fuel ratio AFtrg.

If the determination result in step S18 is negative, the ECU 6 calculates the learning value LAF (step S20) and updates the correction value CAF using this learning value LAF (step S22). The method for calculating the learning value LAF has already been described.

If the determination result in step S18 is affirmative, the ECU 6 determines whether or not the time series value TS has decreased over time (step S24). That is, the ECU 6 determines whether or not the time series value TS fluctuates toward the rich side over time. If the determination result in step S24 is negative, the ECU 6 performs the processes of steps S20 and S22.

If the determination result in step S24 is affirmative, the ECU 6 changes the target air-fuel ratio AFtrg (step S26). Specifically, the ECU 6 sets the average value AVE calculated in step S14 to the new target air-fuel ratio AFtrg.

3. 3. Effect FIG. 7 is a diagram for explaining the effect of executing the learning process according to the present embodiment. Similar to the example shown in FIG. 3, in the example shown in FIG. 7, the true air-fuel ratio AFrl does not match the optimum purification point. Further, as shown in FIG. 7 (i), the rear air-fuel ratio AFrr is in agreement with the target air-fuel ratio AFtrg. Therefore, the problem described in FIG. 3 (i) can also occur in FIG. 7 (i).

Further, as in the example shown in FIG. 4, in the example shown in FIG. 7, the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg due to the deterioration of the three-way catalyst 3. This is the cause of the above problem.

In this regard, according to the learning process according to the present embodiment, it is determined whether or not the optimum purification point fluctuates to the rich side of the target air-fuel ratio AFtrg. Then, when it is determined that the optimum purification point has changed, the target air-fuel ratio AFtrg is changed as shown in FIG. 7 (ii). In this way, when it is determined that the optimum purification point has fluctuated, the learning value LAF is not calculated. Therefore, it is possible to avoid updating the correction value CAL when the optimum purification point fluctuates. Further, by changing the target air-fuel ratio AFtrg, it is possible to improve the accuracy of the sub-feedback control and bring the true air-fuel ratio AFrl closer to the optimum purification point. Therefore, it is possible to purify harmful components by fully utilizing the purifying ability of the three-way catalyst 3.

4. Other Examples of Learning Process In the above embodiment, it was determined whether or not the optimum purification point fluctuated to the rich side of the target air-fuel ratio AFtrg based on the tendency of the time series value TS. In this example, this determination is made by changing the element temperature or the circuit temperature of the air-fuel ratio sensor 5 during the learning process. Increasing the heater power of the air-fuel ratio sensor 5 raises the element temperature or the circuit temperature. When the fan speed of the air-fuel ratio sensor 5 is changed, the element temperature or the circuit temperature rises or falls.

The average value AVE changes before and after the temperature change. When the cause of the deviation between the rear air-fuel ratio AFrr and the true air-fuel ratio AFrr is the air-fuel ratio sensor 5, the deviation of the average value AVE before and after tends to be large. Therefore, in this example, when the deviation between the average value AVE before the change and the average value AVE after the change is larger than the threshold value, it is determined that the cause of the deviation is the air-fuel ratio sensor 5, and the correction value CAF is updated. If not, it is determined that the cause of the deviation is the change of the optimum purification point to the rich side, and the average value AVE before the change is set to the new target air-fuel ratio AFtrg.

According to another example of the learning process described above, it is possible to obtain the same effect as the effect of the learning process according to the above embodiment.

1 Internal combustion engine 2 Exhaust pipe 3 Three-way catalyst 4, 5 Air-fuel ratio sensor 6 ECU
10 Exhaust purification system AFfr Front air-fuel ratio AFrr True air-fuel ratio AFrr Rear air-fuel ratio AFth Theoretical air-fuel ratio AFtrg Target air-fuel ratio AVE Average value CAF Correction value Dfr, Dr, Dave deviation LAF Learning value TS Time series value

Claims (1)

A catalyst that purifies the exhaust gas from the internal combustion engine,
An air-fuel ratio sensor that detects the rear air-fuel ratio, which is the air-fuel ratio of the exhaust gas that has passed through the catalyst, and
A control device that performs air-fuel ratio feedback control based on the deviation between the rear air-fuel ratio and the target air-fuel ratio, and learning processing for learning the correction value of the rear air-fuel ratio.
With
The control device is used in the learning process.
The learning value is calculated based on the deviation,
When the rear air-fuel ratio is a value on the lean side of the target air-fuel ratio, the correction value is updated using the learning value.
When the rear air-fuel ratio is a value on the rich side of the target air-fuel ratio, it is determined whether or not the time series value of the deviation including the current value of the deviation fluctuates to the rich side over time.
When it is determined that the time-series value does not fluctuate toward the rich side over time, the learning value is used to update the correction value.
When it is determined that the time-series value fluctuates toward the rich side over time, the current value of the rear air-fuel ratio is set to a new target air-fuel ratio in the air-fuel ratio feedback control. Engine exhaust purification system.

Images