JP2002364427A - Air-fuel ratio controller for engine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To control an air-fuel ratio with high accuracy and high response by air-fuel ratio learning control based on the air-fuel ratio detected from an output value of an oxygen sensor. SOLUTION: From an air-fuel ratio feedback correction coefficient α and an actual equivalence ratio found from the output value Es of the oxygen sensor, a correction request value is calculated as the correction request value = α/the actual equivalence ratio, and a weighted average result of the correction request value and a previous air-fuel ratio learning correction value is updated and stored as a new air-fuel ratio learning correction value. A weighting factor k used for the weighted average is changed according as the output value Es of the oxygen sensor is in the vicinity of a theoretical air-fuel ratio corresponding value to suppress erroneous learning when detection accuracy of the actual equivalence ratio reduces.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an air-fuel ratio control device for an engine, and more particularly to a technology for detecting an air-fuel ratio based on an output value of an oxygen sensor and learning and correcting an actual air-fuel ratio to a target air-fuel ratio. .

[0002]

2. Description of the Related Art Conventionally, an oxygen sensor whose output value changes in response to oxygen concentration in exhaust gas has been provided, and an actual air-fuel ratio has been obtained by converting the output value of the oxygen sensor into air-fuel ratio data. Feedback control of an air-fuel ratio control signal is performed based on a deviation between an actual air-fuel ratio and a stoichiometric air-fuel ratio as a target air-fuel ratio (Japanese Patent Laid-Open No. 07-127505).
Reference).

[0003] With the above configuration, it is possible to stably converge to the vicinity of the target air-fuel ratio with respect to the control for performing only the rich / lean determination with respect to the stoichiometric air-fuel ratio. Also, in order to make the oxygen concentration in a predetermined vacant room a reference value, energization to the sensor element for transporting oxygen ions is controlled, and the air-fuel ratio is linearly detected from the energized current value (limit current) at that time. Since the oxygen sensor is cheaper than the air-fuel ratio sensor, the system cost can be reduced.

[0004]

In the system using the wide-range air-fuel ratio sensor, the wide-range air-fuel ratio sensor has high detection accuracy in a relatively wide range including the stoichiometric air-fuel ratio. From this and the value of the air-fuel ratio feedback correction coefficient at that time, the air-fuel ratio learning correction value can be updated and learned with high convergence responsiveness.

However, in the case where the output value of the oxygen sensor is converted into data of the air-fuel ratio as described above, in a rich / lean region far away from the vicinity of the stoichiometric air-fuel ratio, the output of the oxygen sensor changes with respect to the change in the air-fuel ratio. Hardly changes, the detection accuracy of the air-fuel ratio is greatly reduced, and even within the range in which the output changes in accordance with the change in the air-fuel ratio, the detection accuracy of the air-fuel ratio decreases as the output departs from near the stoichiometric air-fuel ratio. .

Therefore, if the air-fuel ratio learning is performed in the same manner as in the case where the air-fuel ratio is detected by the wide area air-fuel ratio sensor, there is a problem that the learning accuracy cannot be ensured. The present invention has been made in view of the above problems,
Provided is an air-fuel ratio control device for an engine that can accurately perform an air-fuel ratio learning based on a detection result of an air-fuel ratio in a system that performs air-fuel ratio feedback control by converting an output value of an oxygen sensor into air-fuel ratio data. The purpose is to do.

[0007]

According to the present invention, an oxygen sensor whose output value changes in response to the oxygen concentration in exhaust gas is provided, and the output value of the oxygen sensor is converted into air-fuel ratio data. To detect the actual air-fuel ratio,
While the air-fuel ratio feedback correction value is set based on a comparison between the actual air-fuel ratio and the target air-fuel ratio, when the output value of the oxygen sensor is within a predetermined range including a stoichiometric air-fuel ratio equivalent value, the actual air-fuel ratio is corrected. The air-fuel ratio learning correction value is updated based on the fuel ratio and the air-fuel ratio feedback correction value.

According to this configuration, the actual air-fuel ratio and the air-fuel ratio feedback correction detected in the region near the stoichiometric air-fuel ratio except for the rich / lean region where the detection accuracy of the air-fuel ratio based on the output value of the oxygen sensor is greatly reduced. The air-fuel ratio learning correction value for matching the actual air-fuel ratio to the target air-fuel ratio is updated based on the value. In the invention according to claim 2,
When the output value of the oxygen sensor is continuously included in the predetermined range, the air-fuel ratio learning correction value is updated.

With this configuration, when it is continuously determined that the output value of the oxygen sensor is within the range including the stoichiometric air-fuel ratio equivalent value, it is determined that the state is stable within the range. The air-fuel ratio learning is performed. In the invention according to claim 3, when the output value of the oxygen sensor is included in a narrower range including the stoichiometric air-fuel ratio equivalent value included in the predetermined range, the update speed of the air-fuel ratio learning correction value is changed. The configuration was accelerated.

According to this configuration, the air-fuel ratio learning is performed within a predetermined range including the stoichiometric air-fuel ratio equivalent value. However, the air-fuel ratio learning is performed within a range closer to the stoichiometric air-fuel ratio within the predetermined range and with higher accuracy. When the air-fuel ratio can be detected, the update speed of the air-fuel ratio learning correction value is increased as compared to the case where the air-fuel ratio is an area outside the stoichiometric air-fuel ratio. The update speed indicates how much the latest detected air-fuel ratio correction request is reflected in the air-fuel ratio learning correction value. When the update speed is high, the latest correction request includes the air-fuel ratio learning correction value. Will approach faster.

According to a fourth aspect of the present invention, an oxygen sensor whose output value changes in response to the oxygen concentration in the exhaust gas is provided, and the output value of the oxygen sensor is converted into air-fuel ratio data to convert the actual air-fuel ratio. The air-fuel ratio feedback correction value is set based on the comparison between the actual air-fuel ratio and the target air-fuel ratio, and the air-fuel ratio learning correction is set based on the actual air-fuel ratio and the air-fuel ratio feedback correction value. While the value is updated, the update speed of the air-fuel ratio learning correction value is determined based on which of the ranges the output value of the oxygen sensor falls into around the stoichiometric air-fuel ratio equivalent value. It was configured to change.

According to the above configuration, the actual air-fuel ratio is made to match the target air-fuel ratio from the actual air-fuel ratio detected by converting the output value of the oxygen sensor into air-fuel ratio data and the air-fuel ratio feedback correction value. The update speed at which the air-fuel ratio learning correction value is updated is changed according to whether or not the detected air-fuel ratio at that time is near the stoichiometric air-fuel ratio, in other words, according to the air-fuel ratio detection accuracy.

According to a fifth aspect of the present invention, the updating speed of the air-fuel ratio learning correction value is increased as the output value of the oxygen sensor is included in a range closer to the stoichiometric air-fuel ratio equivalent value. According to this configuration, since the detection accuracy of the air-fuel ratio increases as the output value of the oxygen sensor approaches the stoichiometric air-fuel ratio equivalent value, the update speed of the air-fuel ratio learning correction value increases as the detection accuracy of the air-fuel ratio increases. Become.

According to a sixth aspect of the present invention, there is provided an evaporative fuel processing device that adsorbs and collects the evaporative fuel generated in the fuel tank in a canister and supplies the evaporative fuel desorbed from the canister to an engine. Updating the air-fuel ratio learning correction value based on the air-fuel ratio feedback correction value corrected according to the amount of fuel supplied to the engine by the desorption when the evaporative fuel is being desorbed from the canister. And the update speed of the air-fuel ratio learning correction value is reduced.

According to this configuration, when the learning is performed in anticipation of the rich change in the air-fuel ratio due to the supply of the fuel desorbed from the canister to the engine, the learning accuracy of the air-fuel ratio learning correction value decreases. Therefore, the learning is performed by reducing the update speed. In the invention described in claim 7, the update speed of the air-fuel ratio learning correction value is reduced when the evaporative fuel is desorbed from the canister and within a predetermined time after the desorption is stopped. The configuration was adopted.

According to such a configuration, even after the desorption is stopped, it takes time for all the evaporated fuel supplied to the intake system of the engine to be sucked into the engine and burned.
When it is within a predetermined time after the detachment is stopped, the update speed is reduced as in the case where the detachment is performed.

[0017]

According to the first aspect of the present invention, learning of the air-fuel ratio in a region where the detection accuracy of the air-fuel ratio based on the output value of the oxygen sensor is greatly reduced is avoided, and high convergence is achieved while using an inexpensive oxygen sensor. Thus, there is an effect that the air-fuel ratio learning can be performed with the characteristic. According to the second aspect of the present invention, the air-fuel ratio learning is performed on the condition that the air-fuel ratio detection accuracy is stable in an area where the accuracy can be ensured, so that the air-fuel ratio learning can be stably performed.

According to the third aspect of the present invention, when the detection accuracy of the air-fuel ratio is close to the stoichiometric air-fuel ratio, the update speed of the air-fuel ratio learning correction value is increased so that the air-fuel ratio learning correction value is set to the required value. Has the effect of being able to converge with high accuracy and good response. According to the fourth aspect of the present invention, the air-fuel ratio learning correction value is updated with good convergence response while avoiding erroneous learning by changing the update speed of the air-fuel ratio learning value in correlation with the air-fuel ratio detection accuracy. There is an effect that can be.

According to the present invention, when the air-fuel ratio detection accuracy is close to the stoichiometric air-fuel ratio, the air-fuel ratio learning correction value can be updated with high accuracy and high convergence response. . According to the sixth aspect of the present invention, when the learning accuracy is reduced by the supply of the evaporative fuel desorbed from the canister to the engine, the update speed of the air-fuel ratio learning correction value is reduced, so that the erroneous learning is prevented. There is an effect that the air-fuel ratio learning correction value can be converged while avoiding.

According to the seventh aspect of the present invention, the air-fuel ratio learning correction value is increased at a high update speed within a predetermined time after the desorption is stopped and the effect of the desorbed evaporated fuel remains. There is an effect that erroneous learning can be avoided.

[0021]

Embodiments of the present invention will be described below. FIG. 1 is a system configuration diagram of an engine according to the embodiment. In FIG. 1, air is sucked into the combustion chamber of each cylinder of an engine 1 mounted on a vehicle via an air cleaner 2, an intake passage 3, and an electronically controlled throttle valve 4 driven to open and close by a motor.

An electromagnetic fuel injection valve 5 for directly injecting fuel (gasoline) into the combustion chamber of each cylinder is provided, and the fuel injected from the fuel injection valve 5 and the sucked air enter the combustion chamber. An air-fuel mixture is formed. Fuel injection valve 5
The solenoid is energized by the injection pulse signal output from the control unit 20 to open the valve and injects fuel adjusted to a predetermined pressure.

The injected fuel diffuses into the combustion chamber in the case of the intake stroke injection to form a homogeneous mixture, and in the case of the compression stroke injection, the fuel is concentrated in a layered manner around the ignition plug 6. To form The mixture formed in the combustion chamber is ignited and burned by the ignition plug 6. However, the engine 1 is not limited to the direct injection gasoline engine described above.
The engine may be configured to inject fuel into the intake port.

Exhaust gas from the engine 1 is exhausted from an exhaust passage 7, and an exhaust purification catalyst 8 is interposed in the exhaust passage 7. The catalyst 8 is a three-way catalyst that oxidizes carbon monoxide CO and hydrocarbons HC, which are three harmful components in the exhaust gas, and reduces nitrogen oxides NOx to harmless carbon dioxide, water vapor and nitrogen. Is to be converted.

The purifying performance of the three-way catalyst 8 is as follows:
When the exhaust air-fuel ratio is the stoichiometric air-fuel ratio, and when the exhaust air-fuel ratio is lean and the amount of oxygen is excessive, the oxidizing action becomes active but the reducing action becomes inactive, and conversely, the exhaust air-fuel ratio becomes rich. When the oxygen content is small, the oxidizing action becomes inactive but the reducing action becomes active. Further, an evaporative fuel processing device for performing a combustion process on the evaporative fuel generated in the fuel tank 9 is provided.

The canister 10 is a sealed container filled with an adsorbent 11 such as activated carbon, and is connected to an evaporative fuel introduction pipe 12 extending from the fuel tank 9. Therefore, the evaporated fuel generated in the fuel tank 9 passes through the evaporated fuel introduction pipe 12 and is guided to the canister 10 to be adsorbed and collected. The canister 10 has a fresh air inlet 13.
Is formed, and a purge pipe 14 is led out. A purge control valve 15 whose opening and closing are controlled by a control signal from a control unit 20 is interposed in the purge pipe 14.

In the above configuration, when the purge control valve 15 is controlled to open, the suction negative pressure of the engine 1 is reduced by the canister 10.
As a result, the evaporative fuel adsorbed on the adsorbent 11 of the canister 10 is purged by the air introduced from the fresh air inlet 13, and the purge air passes through the purge pipe 14 to the downstream of the throttle valve 4 in the intake passage 3. It is sucked and then burned in the combustion chamber of the engine 1.

The control unit 20 includes a CPU, R
A microcomputer including an OM, a RAM, an A / D converter, an input / output interface, and the like;
Input signals from various sensors are received, and arithmetic processing is performed based on these signals to obtain throttle valves 4, fuel injection valves 5, ignition plugs 6,
And the operation of the purge control valve 15 and the like. As the various sensors, a crank angle sensor 21 for detecting a crank angle of the engine 1 and a cam sensor 22 for extracting a cylinder discrimination signal from a camshaft are provided. Based on a signal from the crank angle sensor 21, the rotation speed of the engine 1 is reduced. Is calculated.

In addition, an air flow meter 23 for detecting the intake air flow rate Qa upstream of the throttle valve 4 in the intake passage 3,
An accelerator sensor 24 for detecting an accelerator pedal depression amount (accelerator opening) APS, an opening TV of the throttle valve 4
A throttle sensor 25 for detecting O, a water temperature sensor 26 for detecting the cooling water temperature Tw of the engine 1, an oxygen sensor 27 whose output value changes in response to the oxygen concentration in the exhaust gas, a vehicle speed VS
A vehicle speed sensor 28 for detecting P is provided.

The oxygen sensor 27 is disclosed in JP-A-11-32
As disclosed in Japanese Patent Application Laid-Open No. 6266, the zirconia tube has a zirconia tube protrudingly provided in an exhaust pipe. The zirconia tube has an oxygen concentration in the exhaust gas outside the zirconia tube and an oxygen concentration in the air inside the zirconia tube. This is a zirconia oxygen sensor that generates an electromotive force. Output value E of the oxygen sensor 27
As shown in FIG. 2, s (electromotive force) changes suddenly at the stoichiometric air-fuel ratio, and the electromotive force is higher on the rich side than the stoichiometric air-fuel ratio.
Although the electromotive force is lower on the lean side than the stoichiometric air-fuel ratio, the protective layer, the catalyst layer, and the protective layer constituting the sensor element are arranged so that the output value changes according to the air-fuel ratio near the stoichiometric air-fuel ratio.
A zirconia tube is formed.

When a predetermined air-fuel ratio feedback control condition is satisfied, the control unit 20 controls the air-fuel ratio feedback correction coefficient (air-fuel ratio) based on the output value of the oxygen sensor 27 so that the actual air-fuel ratio matches the target air-fuel ratio. A fuel-fuel ratio feedback correction value) is calculated, and an air-fuel ratio learning correction value for absorbing the air-fuel ratio variation for each operating condition is learned. The details of the air-fuel ratio control are shown in FIGS. This will be described according to the flowchart.

The flowchart of FIG. 3 shows a main routine. First, in step S1, the output value Es of the oxygen sensor 27, the cooling water temperature Tw, the engine speed Ne, the intake air amount Q, and the like are read. Step S
In 2, it is determined whether or not the execution permission condition of the air-fuel ratio feedback control is satisfied.

Here, it is determined that the oxygen sensor 27 is in an active state as the permission condition. When the execution permission condition of the air-fuel ratio feedback control is satisfied, the process proceeds to step S3, and the air-fuel ratio feedback correction coefficient α is calculated. The details of the air-fuel ratio feedback control will be described with reference to the flowchart of FIG.

In step S301, the oxygen sensor 2
It is determined whether or not the output value Es7 is within a predetermined range.
The predetermined range is a range including a value corresponding to the stoichiometric air-fuel ratio of the sensor output, and is a region where the output value Es relatively changes abruptly with a change in the air-fuel ratio. In other words, the range is richer than the stoichiometric air-fuel ratio. Or, it is a region near the stoichiometric air-fuel ratio except for a region that is lean and the output value Es hardly changes in response to a change in the air-fuel ratio.

In this embodiment using the oxygen sensor 27 having an output characteristic as shown in FIG. 2, the value corresponding to the stoichiometric air-fuel ratio is substantially equal.
0.5 (V), and the predetermined range is 0.3 (V) ≦ Es ≦
0.8 (V). Step S301
When it is determined that the output value Es of the oxygen sensor 27 is within the predetermined range, the process proceeds to step S302, and a process of converting the output value Es of the oxygen sensor 27 into data of the air-fuel ratio is performed.

The conversion may be performed based on a table showing the correlation between the output value Es and the air-fuel ratio. However, in order to further increase the resolution, the output value Es is converted into another value based on a preset calculation formula. After the replacement with the variable, the air-fuel ratio data may be obtained from the variable. Step S3
In 03, a deviation err (deviation err = actual air-fuel ratio-target air-fuel ratio) between the actual air-fuel ratio obtained from the output value Es and the stoichiometric air-fuel ratio as the target air-fuel ratio is obtained.

In step S304, the deviation err is multiplied by a proportional constant Kp to obtain a proportional manipulated variable P (P = err × Kp).
Is calculated. In step S305, by determining whether the deviation err is positive or negative, rich / lean of the actual air-fuel ratio with respect to the stoichiometric air-fuel ratio is determined. When the deviation err is negative and the actual air-fuel ratio is richer than the stoichiometric air-fuel ratio, Step S3
Proceed to 06.

In step S306, the result obtained by subtracting a predetermined value ΔI from the previous value of the integral manipulated variable I is defined as the current integral manipulated variable I. Also, in step S305, the deviation err
Is positive and the actual air-fuel ratio is leaner than the stoichiometric air-fuel ratio, the process proceeds to step S307, and the result of adding the predetermined value ΔI to the previous value of the integral operation amount I is set as the current integral operation amount I. .

Further, when it is determined in step S305 that the deviation err is substantially zero and the actual air-fuel ratio is substantially equal to the stoichiometric air-fuel ratio, the process bypasses steps S306 and 307 and proceeds to step S312, whereby the integration is performed. The operation amount I is held at the previous value. In step S312, the air-fuel ratio feedback correction coefficient α (air-fuel ratio feedback correction value) is calculated as α = P + I + 1.0.

On the other hand, in step S301, the output value Es of the oxygen sensor 27 is out of the predetermined range (0.3 (V)>
If it is determined that Es or Es> 0.8 (V), the process proceeds to step S308. In step S308, it is determined whether or not the output value Es of the oxygen sensor 27 is out of a range higher than the predetermined range, specifically, whether or not Es> 0.8 (V). It is determined whether or not the actual air-fuel ratio is richer than the stoichiometric air-fuel ratio.

In step S308, when it is determined that Es> 0.8 (V), and the actual air-fuel ratio is richer than the stoichiometric air-fuel ratio, the process proceeds to step S309, where the integral manipulated variable I
The result obtained by subtracting the predetermined value ΔI from the previous value of is used as the current integral manipulated variable I. On the other hand, in step S308, Es> 0.8
(V), it is determined that 0.3 (V)> E
s, so that steps S301 to S308
It is determined that the actual air-fuel ratio is leaner than the stoichiometric air-fuel ratio.

In this case, the routine proceeds to step 310, where the result obtained by adding the predetermined value ΔI to the previous value of the integral manipulated variable I is defined as the current integral manipulated variable I. After setting the integral manipulated variable I in steps S309 and S310, 0 is set in the proportional manipulated variable P in the next step S311. The output value Es of the oxygen sensor 27 is within the above-mentioned predetermined range (0.3 (V) ≦ Es ≦ 0.8
(V)), the output value Es can be accurately converted to the air-fuel ratio data. However, if the output value Es is out of the predetermined range, the output value Es hardly changes with respect to the change in the air-fuel ratio, and the air-fuel ratio is correctly obtained. Therefore, the proportional control based on the air-fuel ratio deviation is prohibited, and the air-fuel ratio is prevented from being controlled based on incorrect information on the air-fuel ratio deviation.

However, since rich / lean discrimination with respect to the stoichiometric air-fuel ratio can be performed, the integral control based on rich / lean discrimination is performed in the same manner as when the output value Es is within a predetermined range. Therefore, when the process proceeds to step S312 when the output value Es is out of the predetermined range, substantially,
The air-fuel ratio feedback correction coefficient α is calculated as α = I + 1.0.

In step S3 of the flowchart of FIG.
When the air-fuel ratio feedback correction coefficient α is calculated as described above, in step S4, the permission of the air-fuel ratio learning is determined as shown in the flowchart of FIG. In the flowchart of FIG. 5, in step S401, it is determined whether or not the output value Es of the oxygen sensor 27 is within a predetermined range.

The predetermined range is the same as the range determined in step S301 (0.3 (V) ≦ Es ≦ 0.8 (V)). If the predetermined range is not within this range, the air-fuel ratio cannot be obtained correctly, and In order to stop the learning using the detection result of the fuel ratio, the process proceeds to step S402, and the permission determination is determined as NG. On the other hand, in step S401, 0.3 (V)
If it is determined that ≦ Es ≦ 0.8 (V), the process proceeds to step S403, and the previous time is also 0.3 (V) ≦ Es ≦ 0.8.
It is determined whether or not (V) has been determined.

0.3 (V) ≦ Es ≦ 0.8 (V) two consecutive times
If it is, it is determined that the air-fuel ratio is stable in an area where the air-fuel ratio can be detected with high accuracy, and the process proceeds to step S405, and the permission determination is made as O.
Let it be K. On the other hand, in step S401, the current output value E
Even if it is determined that s is 0.3 (V) ≦ Es ≦ 0.8 (V), if it is determined in step S403 that the previous value is not within the predetermined range, the process proceeds to step S404, and the permission determination is determined as NG. .

In step S5 of the flowchart of FIG. 3, OK / NG of the permission determination is determined. When the permission determination is OK and the air-fuel ratio learning using the air-fuel ratio detected based on the output of the oxygen sensor 27 can be performed, the process proceeds to step S6, and the update learning of the air-fuel ratio learning correction value is performed.

On the other hand, when the permission judgment is NG, the learning process of step S6 is bypassed, and this routine ends. The details of the air-fuel ratio learning in step S6 are described below.
This will be described with reference to the flowchart of FIG. Step S6
In 01, it is determined whether or not a predetermined time has elapsed since the purge control valve 15 was controlled to be fully closed (opening degree = 0).

When the purge control valve 15 is closed, fuel desorbed from the canister 10 is not newly supplied to the intake passage 3, but is supplied to the intake passage 3 immediately before the purge control valve 15 is closed. It takes a predetermined time for all the fuel that has been burned to undergo the effect of desorption after being burned. In other words, when a predetermined time or more has elapsed since the purge control valve 15 was closed, it can be determined that there is no influence of detachment from the canister 10.

If the predetermined time has elapsed since the purge control valve 15 was closed, the process proceeds to step S602, where the predetermined range determined in steps S301 and S401 (0.3 (V) ≦ Es ≦ 0.8 ( V)), it is determined whether or not the output value Es is included in a range (for example, 0.45 (V) ≦ Es ≦ 0.65 (V)) narrower than that. Step S
The predetermined range (0.3 (V) ≦ E determined at 301, 401)
If s ≦ 0.8 (V), the air-fuel ratio can be obtained from the output value Es of the oxygen sensor 27, but in a region closer to the stoichiometric air-fuel ratio (0.45 (V) ≦ Es ≦ 0.65 (V)). ,
The air-fuel ratio can be detected with higher accuracy.

In step S602, 0.45 (V) ≦ Es ≦
If it is determined to be 0.65 (V), the process advances to step S603 to set k1 (for example, k1 = 0.5) as the weight k used for the weighted average of the air-fuel ratio learning correction values. The weighting weight k is used when a weighted average of the air-fuel ratio learning correction value up to the previous time and the latest calculated correction request is used, and the result is updated and stored as a new air-fuel ratio learning correction value. As the value increases, a value closer to the latest calculated correction request is calculated as the air-fuel ratio learning correction value. In other words, as the weighting weight k increases,
The update speed of the air-fuel ratio learning correction value increases.

On the other hand, in step S602, 0.45 (V) ≦
If it is determined that Es ≦ 0.65 (V) is not satisfied, step S
Proceeding to 604, the weighted weight k used in the weighted average of the air-fuel ratio learning correction value is k2 (k2 <k1, for example, k2 = 0.
25) Set. When 0.45 (V) ≦ Es ≦ 0.65 (V) is not satisfied, the air-fuel ratio can be detected based on the output value Es of the oxygen sensor 27, but 0.45 (V) ≦ Es ≦ 0.
The detection accuracy of the air-fuel ratio is lower than when it is 65 (V).

Therefore, 0.45 (V) ≦ Es ≦ 0.65 (V)
If not, the update speed of the air-fuel ratio learning correction value is reduced so that even if an erroneous correction request is calculated from the air-fuel ratio detection result, it is difficult to reflect the correction request on the air-fuel ratio learning correction value. When the weighting weight k is set in step S603 or step S604, in step S605, based on the air-fuel ratio feedback correction coefficient α and the actual air-fuel ratio (actual equivalent ratio) obtained from the output value Es of the oxygen sensor 27, The correction request value is calculated as correction request value = air-fuel ratio feedback correction coefficient α / actual equivalent ratio, and a weighted average operation of the correction request value and the air-fuel ratio learning correction value up to the previous time is calculated by the weighting k. And the result is updated and stored as a new air-fuel ratio learning correction value.

Air-fuel ratio learning correction value = correction request value × k + previous value × (1-k) As described above, the correction request value is calculated as correction request value = air-fuel ratio feedback correction coefficient α / actual equivalent ratio. With the configuration, the air-fuel ratio learning can be performed without waiting for the convergence of the air-fuel ratio feedback correction coefficient α, but the actual equivalent ratio used for calculating the correction request value is not always detected with the same accuracy. The closer to, the higher the accuracy.

Therefore, by changing the weighting k used in the weighted average calculation of the air-fuel ratio learning correction value in correlation with the detection accuracy, the correction accuracy at the time when the air-fuel ratio learning value converges can be secured. Learning can be advanced from a state apart from the target air-fuel ratio (the stoichiometric air-fuel ratio) to control the target air-fuel ratio to near the target air-fuel ratio at an early stage. When the air-fuel ratio learning correction value is learned for each engine operating condition, the air-fuel ratio learning correction value stored corresponding to the operating condition at that time is read, and the weighted average calculation is performed. Then, the calculation result is updated and stored as a value suitable for the operating condition at that time.

On the other hand, if it is determined in step S601 that the predetermined time has not elapsed since the purge control valve 15 was closed, that is, there is an effect due to the desorption of the fuel vapor from the canister 10. In some cases, the process proceeds to step S606. In step S606, the output value Es of the oxygen sensor 27 is determined in the same manner as in step S602.
Is in a narrow range near the stoichiometric air-fuel ratio (0.45 (V) ≦ Es ≦ 0.65
(V)) It is determined whether it is within.

Here, if 0.45 (V) ≦ Es ≦ 0.65 (V), the air-fuel ratio can be detected with high accuracy.
There is an effect of the desorbed fuel from the canister 10, and as described later, even if a correction for estimating the desorbed fuel is performed, the accuracy of the air-fuel ratio learning is lower than in the state where the desorbed fuel is not affected. descend. Therefore, in step S606, 0.45
If it is determined that (V) ≦ Es ≦ 0.65 (V), the process proceeds to step S607, and k3 (for example, k3 = 0.25) smaller than k1 is set as the weighting weight k.

In step S606, 0.45 (V) .ltoreq.E
When it is determined that s ≦ 0.65 (V) is not satisfied, there is an effect of the desorbed fuel, and the condition is that the accuracy of detecting the air-fuel ratio by the oxygen sensor 27 is reduced. The smallest value k4 (for example, k4 = 0.
Set 125). In step S609, a correction request value is calculated according to the following equation.

Correction required value = (α- (purge flow coefficient × boost × purge control signal / fuel injection amount)) / actual equivalent ratio The purge flow coefficient is preset according to the purge pipe 14 and the purge control valve 15. The purge control signal is a control signal of the purge control valve 15 and is a value indicating the opening degree at that time.

As described above, the influence of the fuel desorbed from the canister 10 is used to calculate the air-fuel ratio feedback correction coefficient α.
Is subtracted from the above to calculate the correction request value.
07 or the weighting weight k set in step S608, the update calculation of the air-fuel ratio learning correction value is performed according to the following equation. Air-fuel ratio learning correction value = correction request value × k + previous value × (1-
k) The basic fuel injection amount Tp is multiplied by the air-fuel ratio feedback correction coefficient α and the air-fuel ratio learning correction value in the calculation of the fuel injection amount Ti.

Ti = Tp × α × Air-fuel ratio learning correction value × CO
+ Ts In the above, CO is various correction coefficients calculated based on the cooling water temperature and the like, and Ts is a correction component based on the voltage of the battery that is the power source of the fuel injection valve 5.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram of an engine according to an embodiment.

FIG. 2 is an output characteristic diagram of the oxygen sensor according to the embodiment.

FIG. 3 is a flowchart showing a main routine of air-fuel ratio control in the embodiment.

FIG. 4 is a flowchart showing details of air-fuel ratio feedback control in the embodiment.

FIG. 5 is a flowchart showing details of air-fuel ratio learning permission determination in the embodiment.

FIG. 6 is a flowchart showing details of air-fuel ratio learning control in the embodiment.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Engine 4 ... Throttle valve 5 ... Fuel injection valve 6 ... Spark plug 8 ... Catalyst 10 ... Canister 14 ... Purge piping 15 ... Purge control valve 20 ... Control unit 21 ... Crank angle sensor 23 ... Air flow meter 27 ... Oxygen sensor

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 3G084 BA09 BA13 DA05 EA02 EA11 EB14 EB15 EB20 FA07 FA20 FA29 FA33 3G301 HA01 HA04 HA06 HA14 JA21 LB04 LC03 MA01 MA11 NA03 NA04 NA09 ND05 ND27 PA01Z PB09Z PD03Z PE01Z PE08Z

Claims (7)

[Claims] An oxygen sensor whose output value changes in response to the oxygen concentration in exhaust gas is provided, and an output value of the oxygen sensor is converted into air-fuel ratio data to detect an actual air-fuel ratio. While the air-fuel ratio feedback correction value is set based on the comparison between the actual air-fuel ratio and the target air-fuel ratio, when the output value of the oxygen sensor is within a predetermined range including a stoichiometric air-fuel ratio equivalent value, the actual air-fuel ratio is corrected. An air-fuel ratio control device for an engine configured to update an air-fuel ratio learning correction value based on a fuel ratio and the air-fuel ratio feedback correction value. 2. The engine according to claim 1, wherein the air-fuel ratio learning correction value is updated when an output value of the oxygen sensor is continuously included in the predetermined range. Fuel ratio control device. 3. When the output value of the oxygen sensor is included in a narrower range including the stoichiometric air-fuel ratio equivalent value included in the predetermined range, the updating speed of the air-fuel ratio learning correction value is increased. The air-fuel ratio control device for an engine according to claim 1 or 2, wherein: 4. An oxygen sensor whose output value changes in response to the oxygen concentration in the exhaust gas, wherein the output value of the oxygen sensor is converted into air-fuel ratio data to detect an actual air-fuel ratio. An air-fuel ratio feedback correction value is set based on a comparison between the actual air-fuel ratio and a target air-fuel ratio, and an air-fuel ratio learning correction value is updated based on the actual air-fuel ratio and the air-fuel ratio feedback correction value. On the other hand, the configuration is such that the update speed of the air-fuel ratio learning correction value is changed depending on which of the ranges divided by the output value of the oxygen sensor around the stoichiometric air-fuel ratio equivalent value is the center. An air-fuel ratio control device for an engine. 5. The engine according to claim 4, wherein the update speed of the air-fuel ratio learning correction value is increased as the output value of the oxygen sensor is included in a range closer to the stoichiometric air-fuel ratio equivalent value. Air-fuel ratio control device. 6. An evaporative fuel treatment device which adsorbs and collects the evaporative fuel generated in the fuel tank to a canister and supplies the evaporative fuel desorbed from the canister to an engine, wherein the evaporative fuel from the canister is provided. While the fuel is being desorbed, the air-fuel ratio learning correction value is updated based on the air-fuel ratio feedback correction value corrected in accordance with the amount of fuel supplied to the engine by the desorption, and The air-fuel ratio control device for an engine according to any one of claims 1 to 5, wherein an update speed of the air-fuel ratio learning correction value is reduced. 7. A method for reducing the update speed of the air-fuel ratio learning correction value when the vaporized fuel is being desorbed from the canister and within a predetermined time after the desorption is stopped. 7. An air-fuel ratio control device for an engine according to claim 6, wherein: