KR930007609B1 - Air-fuel ratio control system for internal combustion engine - Google Patents

Abstract

No content.

Description

Air-fuel ratio control device of internal combustion engine

1 to 8 show an air-fuel ratio control apparatus for an internal combustion engine as a first embodiment of the present invention, and FIG. 1a shows a control block diagram thereof.

Fig. 1B is a main part control block diagram.

2 is a block diagram mainly showing the hardware.

3 is an overall configuration diagram showing the engine system.

4A to 4D are flowcharts for explaining any of the juroutines.

5 is a flowchart for explaining the solenoid valve driving routine.

6A is a flowchart for explaining the timer subtraction routine.

6B is a flowchart for explaining the integral time calculation routine.

7A to 7C are graphs for explaining the air-fuel ratio feedback correction coefficients.

8A to 8C are graphs for explaining the action.

9 to 12 are control block diagrams showing an air-fuel ratio control apparatus for an internal combustion engine as a second embodiment of the present invention.

9 is the main control block diagram.

10A to 10E are flowcharts for explaining the juroutine.

11A to 11C are graphs for explaining the air-fuel ratio feedback correction coefficient.

12A to 12B are graphs for explaining the response time of the O ₂ detector.

FIG. 13 is a main part control block diagram showing an air-fuel ratio control apparatus for the internal combustion engine of FIGS. 13 to 31 as a third embodiment of the present invention.

14A to 14D are flow charts for explaining the main routine.

Fig. 15 is a flowchart for calculating a deviation value between the downstream O ₂ detector output and a target value.

FIG. 16 is a flow chart for guaranteeing response delay time based on the deviation value obtained in FIG. 15. FIG.

FIG. 17 is a flowchart for correcting the air-fuel ratio feedback gain based on the deviation value obtained in FIG. 15. FIG.

18 is a flowchart for correcting the air-fuel ratio feedback proportional gain based on the deviation value found in FIG.

19A to 20B are graphs for explaining the response delay time correction amount.

21A to 22B are graphs for the air-fuel ratio feedback gain correction amount.

23A to 24B are graphs for the air-fuel ratio feedback proportional gain correction amount.

25A to 26C are graphs for explaining a correction method based on response delay time.

27A to 28C are graphs for explaining the correction method by the air-fuel ratio feedback integral gain.

29A to 30C are graphs for explaining the correction method by the air-fuel ratio feedback proportional gain.

31 is a graph showing the V _f -V _f characteristics.

32 is a main part control block diagram showing an air-fuel ratio control apparatus for the internal combustion engine of FIGS. 32 to 53 as a fourth embodiment of the present invention.

33A-33D are flow charts for describing the juroutine.

34 is a flowchart for calculating a deviation value between the downstream O ₂ sensor output and a target value (reference value).

FIG. 35 is a flowchart for correcting a response lag time based on the deviation value obtained in FIG. 34. FIG.

36 is a flowchart for correcting the air-fuel ratio feedback gain based on the deviation value obtained in FIG. 34. FIG.

FIG. 37 is a flowchart for correcting the air-fuel ratio feedback proportional gain based on the deviation value found in FIG. 34. FIG.

38 is a flowchart for correcting a reference value for rich or lean determination to be compared with an upstream O ₂ detector output based on the deviation value found in FIG. 34. FIG.

39A to 40B are graphs for explaining the response delay time correction amount.

41A to 42B are graphs for explaining the air-fuel ratio feedback gain correction amount.

43A to 44B are graphs for explaining the air-performance ratio feedback proportional gain correction amount.

45A to 45B are graphs for explaining the correction amount of the reference value for determining the rich or lean to be compared with the upstream O ₂ sensor output.

46A to 47C are graphs for explaining a correction method based on response delay time.

48A to 49C are graphs for explaining the correction method by the air-fuel ratio feedback integral gain.

50A to 51C are graphs for explaining the correction method by the air-fuel ratio feedback proportional gain.

52A to 53C are graphs for explaining a correction method on a reference value for determining rich or lean to be compared with its upstream O ₂ detector output.

54 is an overall configuration diagram showing an engine system having an air-fuel ratio control apparatus for an internal combustion engine as a fifth embodiment of the present invention.

55 and 56 are flowcharts showing modifications of the third and fourth embodiments.

* Explanation of symbols for main parts of the drawings

1: combustion chamber 2: intake passage

3: exhaust passage 4: intake valve

5: ventilation valve 6: air cleaner air purifier

7: throttle valve 8: solenoid valve

9: catalytic converter 10: ISC motor

11 Air Pro Sensor Air Flow Detector 12 Intake Air Temperature Detector

13: atmospheric pressure sensor 14: throttle detector

15: flow switch 16: motor position sensor

17: Upstream O ₂ detector as first oxygen concentration sensor

18: downstream O ₂ detector as second oxygen concentration detector

19: water temperature sensor 20: vehicle speed sensor

21: crank angle detector 22: TDC detector

23: electronic control unit (ECU) 24: storage battery

25: battery sensor 26: ignition switch (key switch)

27: CPU 28,29: input interface

30: A / D converter 31: ROM

32: RAM 33: Battery Backup RAM (BURAM)

34: driver 35: basic driving time determining means

36: air performance compensation means 37: O ₂ detector feedback negative means

38,39: switching means 40: cooling water temperature correction means

41: intake air temperature correction means 42: atmospheric pressure correction means

43: acceleration weight correction means 44: dead time correction means

45, 45 ': Rich and lean judgment voltage setting means 45'A: Air-fuel ratio control correction means

46: comparison means 47, 47 ', 47 ": correction coefficient determining means

47'A: Air-fuel ratio control correction means 48: Standard value change means

49: characteristic calculation means

49A: reference value setting means for the downstream O ₂ detector as a reference value setting means for the second oxygen concentration sensor

50: reference value changing means 51: reference value setting means

60: secondary air supply passage 61: electronic control valve

E: organ F: air filter

P: Electric Pump

The present invention uses the detection signal from an oxygen concentration detector (hereinafter referred to as an "O ₂ detector") installed in an exhaust machine of an internal combustion engine (hereinafter referred to as an "engine") as a feedback signal, thereby providing an air-fuel ratio of the internal combustion engine. It relates to the air-fuel ratio control device of the internal combustion engine to control the.

Conventionally, various air-fuel ratio control apparatuses for such internal combustion engines have been proposed, but in this type of air-fuel ratio control apparatus for internal combustion engines, a sensor is fabricated by applying a principle of an oxygen shade cell of a solid electrolyte so that its output value changes rapidly around the theoretical air-fuel ratio. It is located upstream from the catalytic converter (three-way catalyst) excretion part in the machine, and the output from this sensor is the required reference value (this reference value is given as the middle value of the rapidly changing value as a fixed value. The air-fuel ratio of the internal combustion engine is controlled by controlling the fuel injection amount from the electronic fuel injection valve (injector) so that the sensor output is leaned above this reference value and vice versa.

In addition, (hereinafter referred to as the upstream O ₂ sensor to the O ₂ sensor) in the O ₂ sensor downstream portion of the catalytic converter installed in the recent engine ship machinery for installation, and the auxiliary information of the air-fuel ratio control the output from the downstream O ₂ sensor It has also been proposed to be used as a so-called dual O ₂ detector system or a double O ₂ detector system, but in this case, the reference value to be compared with the output of the downstream detector is not changed once set.

As such words for the O ₂ sensor In In the air-fuel ratio control apparatus for a conventional combustion engine of the electronic device to output it performs air-fuel ratio feedback control, and the room for the improved control accuracy also latter downstream O ₂ sensor In the device only by the upstream sensor Since the reference value is a fixed value, air-fuel ratio feedback control based on the output of the upstream O ₂ detector may not be well performed, which may also be improved.

In addition, in the conventional air-fuel ratio control apparatus of the internal combustion engine, since the reference value to be compared with the output of the upstream O ₂ detector is a fixed value in all cases, the control accuracy is not limited by the variation of the characteristics of each product of the O ₂ detector or the secular variation of the characteristics. In addition, there is a problem in terms of control reliability because the exhaust gas purification efficiency by the catalytic converter is also varied.

The present invention is directed to solving these problems the upstream O ₂ sensor and the downstream O ₂ downstream O ₂ sensor is installed the reference value to be compared with the output from one of O ₂ detectors in the detector within or downstream of the upstream sensor and the catalytic converter The control accuracy is not changed by nonuniformity of characteristics or secular variation of characteristics of each product of O ₂ detector, and it is possible to maintain high exhaust gas purification efficiency by catalytic converter. It is an object of the present invention to provide an air-fuel ratio control device for an internal combustion engine that achieves reliability.

In addition, the present invention can change, based on each output of the upstream O ₂ sensor and the downstream O ₂ second upstream O ₂ sensor the reference value to be compared with the output of the other of the O ₂ sensor in the detector and the downstream O ₂ sensor It is an object of the present invention to provide an air-fuel ratio control device for an internal combustion engine in which high control reliability is obtained.

In order to achieve the above object, the air-fuel ratio control apparatus of the internal combustion engine according to the present invention is provided on the upstream side of the catalytic converter for exhaust gas purification installed in the exhaust system of the internal combustion engine to detect the oxygen concentration in the exhaust gas. one side of the _second detector and the catalytic converter of the internal or be installed in the downstream O ₂ sensor and an O ₂ sensor in the second the first the O ₂ sensor of claim 2 for detecting the oxygen concentration in the exhaust gas and at the same time matching Air-fuel ratio control means for controlling the air-fuel ratio of the internal combustion engine based on a comparison result between the detection value from the detector and the required reference value, and based on the outputs from the first O ₂ detector and the second O ₂ detector. It characterized in that the reference value changing means for changing the reference value is provided.

Moreover, the air-fuel ratio control apparatus of the internal combustion engine described in Claim (2) which concerns on this invention is installed in the upstream of the exhaust gas purification catalyst converter provided upstream of the exhaust gas purification catalyst converter installed in the exhaust machine of an internal combustion engine, in addition to the O ₂ sensor of claim 2 for detecting the oxygen concentration in the gas and matching comparison with a reference value of the first of the O ₂ sensor and a second of the O ₂ sensor detection value and the time taken from one O ₂ detectors in the by reference to the results gatchumyeo the air-fuel ratio control means for controlling the air-fuel ratio of the internal combustion engine, O ₂ sensor based on the first of the above by reference to an output from the O ₂ sensor and the second O ₂ sensor in the first of the second of O of the second reference value setting means and the same first of the second reference value and the set by a second reference value setting means for _second detector sets the second reference value for the O ₂ sensor on the other side of the O ₂ sensor and a second O ₂ sensor by reference to the comparison result of the output from the other side of the O ₂ sensor in the and is characterized in that the air-fuel ratio control correction means for applying a correction for the air-fuel ratio control by the air-fuel ratio control means is provided.

The standard value of the detected value and the time taken from the internal combustion engine air-fuel ratio control by the action of the air-fuel ratio control means In the device of claim 1 of the O ₂ sensor and the 2 O ₂ detectors side O ₂ detectors of the invention described in claim 1 of the above-described comparison and air-fuel ratio control of the internal combustion engine, etc. to control a fuel injection amount, based on the comparison result, but the reference value is output from the O ₂ sensor and an O ₂ sensor in the second of the first under a condition that takes a reference value changing means It is changed and updated based on.

After this update, the air-fuel ratio of the internal combustion engine is controlled based on a result of comparing this updated new reference value with the output from the first O ₂ detector.

In addition, in the air-fuel ratio control apparatus of the internal combustion engine described in claim (2) above, the detection value and the required reference value from one of the first O ₂ detector and the second O ₂ detector are caused by the action of the air-fuel ratio control means. wareul compare and the like by controlling a fuel injection amount, based on the comparison result there is also an internal combustion engine air-fuel ratio control based on the output from the 1 O ₂ sensor and the second O ₂ sensor of the second reference value setting means to the second reference value is set, and by the second reference value and the O ₂ sensor and an air-fuel ratio control means for correcting, based on the comparison result of the output from the other side of the O ₂ sensor in the O ₂ sensor 2 of the first Correction is added to the air-fuel ratio control by the air-fuel ratio control means.

1 to 8 show an air-fuel ratio control apparatus for an internal combustion engine as a first embodiment of the present invention. FIG. 1a is a control block diagram and FIG. 1b is a main part thereof. FIG. 2 is a block diagram mainly showing the hardware thereof, FIG. 3 is an overall configuration diagram showing the engine system, and FIGs. 4A to 4D are flowcharts for explaining the main routines; 5 is a flowchart for explaining the solenoid valve driving routine, FIG. 6a is a flowchart for explaining the timer subtraction routine, FIG. 6b is a flowchart for explaining the integral time calculation routine, and FIG. 7 is an air-fuel ratio feedback correction coefficient. 8A to 8C are graphs for explaining the operation thereof.

By the way, the engine system controlled by this apparatus is shown in FIG. 3, in which the engine E has an intake passage 2 and an exhaust passage 3 which communicate with the combustion chamber 1, and the intake passage 2 And the combustion chamber 1 are controlled by the intake valve 4, and the exhaust passage 3 and the combustion chamber 1 are controlled by the exhaust valve 5.

In addition, the air intake passage 2 is provided with an air cleaner 6, a throttle valve 7, and an electronic fuel injection valve (electromagnetic valve) 8 in order from the upstream side, and the exhaust passage 3 is provided in turn from the upstream side. A catalytic converter (three-way catalyst) 9 for exhaust gas purification and a muffler (silencer) not shown are provided.

In addition, the solenoid valve 8 is provided in the intake manifold part by the number of cylinders. If the engine E of the embodiment is a four-cylinder engine in series, four solenoid valves 8 are provided. In other words, the engine may be a so-called multi-point fuel injection (MPI) system.

In addition, the throttle valve 7 is connected to the axel pedal via a wire cable, and the opening degree changes according to the amount of operation of the axel pedal. However, the throttle valve 7 is also opened and closed by an idle speed control motor (ISC motor) 10. This makes it possible to change the opening degree of the throttle valve 7 even when the accelerator pedal is not operated at idle.

In this configuration, the air sucked through the air purifier 6 in accordance with the opening of the throttle valve 7 is mixed with the fuel from the solenoid valve 8 at the intake manifold so as to have an appropriate air-fuel ratio and ignited in the combustion chamber 1. By igniting the plug at the appropriate timing, it is combusted to generate engine torque, and then the mixer is discharged into the exhaust passage 3 as exhaust gas, and the catalytic converter 9 contains three harmful components of Co, Hc and Nox in the mag gas. After being cleaned, it is muffled and released to the atmosphere.

Moreover, in order to control this engine E, various sensors are provided. At the intake passage 2 side, an air flow detector 11 for discharging the amount of intake air from the Kalman and the information to the air purifier excretion portion, an intake temperature detector 12 for detecting the intake air temperature, and an atmospheric pressure detector 13 for detecting the atmospheric pressure ), The throttle detector 14, which is a potentiometric meter for detecting the opening degree of the throttle valve 7, the idle switch 15 for detecting the idle state, and the position of the ISC motor 10 The motor position sensor 16 which detects is provided.

In addition, an upstream detector 17 serving as a first oxygen concentration detector for detecting oxygen concentration (O ₂ concentration) in the exhaust gas is provided on the upstream side of the catalytic converter 9 on the exhaust passage 3 side. In the downstream portion of the catalytic converter 9, a downstream O ₂ detector 18 as a second oxygen concentration detection for detecting the O ₂ concentration in the exhaust gas is similarly provided. Here, both the upstream O ₂ detector 17 and the downstream O ₂ detector 18 apply the principle of an oxygen shade cell, which is both a solid electrolyte, and its output voltage has a characteristic of rapidly changing near the theoretical air-fuel ratio. The voltage on the lean side is low and the voltage on the rich side is higher than the theoretical air-fuel ratio.

In addition, the downstream O ₂ detector 18 may be provided inside the catalytic converter 9.

As other sensors, a water temperature sensor 19 for detecting engine cooling water temperature and a vehicle speed detector 20 (see FIG. 2) for detecting a vehicle speed are provided, as shown in FIGS. 1A and 2. Crank angle detector 21 for detecting crank angle (this crank angle detector 21 also serves as a rotation speed detector for detecting engine speed) and TDC detector for detecting top dead center of the first cylinder (reference cylinder) ( 22 are each installed in the distributor.

The detection signals from these detectors 11 to 22 are input to the electronic control unit (ECU) 23.

The ECU 23 also receives a voltage signal from the storage battery sensor 25 that detects the voltage of the storage battery 24 and a signal from the ignition switch (key switch) 26.

The hardware configuration of the ECU 23 is as shown in FIG. 2, but the ECU 23 has a CPU 27 as its main part, and the CPU 27 includes an intake air temperature detector 12, an atmospheric pressure detector 13, Detection signals from the throttle detector 14, the upstream O ₂ detector 17, the downstream O ₂ detector 18 and the battery detector 25 are input via the input interface 28 and the A / D converter 30 and idle. Detection signals from the detector 15, the vehicle speed detector 20 and the ignition switch 26 are input via an input interface 29 to the air flow detector 11, the crank angle detector 21 and the TDC detector 22. The detection signal from is directly input to the input port.

The CPU 27 is connected to a storage battery 24 by a ROM 31 which stores program data and fixed value data via a bus line, a RAM 32 and a storage battery 24 which are sequentially updated. During this time, data is transferred to and from the battery backup RAM (BURAM) 33 which has been backed up by retaining the stored contents.

The CPU 27 is connected to a storage battery 24 by a ROM 31 which stores program data and fixed value data via a bus line, a RAM 32 and a storage battery 24 which are sequentially updated. The contents of the memory are retained and backed up so that data can be exchanged with the battery backup RAM (BURAM) 33.

In addition, data in the RAM 32 is turned off and reset when the ignition switch 26 is turned off.

Now, if only fuel injection control (air fuel ratio control) is taken into consideration, the fuel injection control signal calculated by the following numerical method from the CPU 27 is output via the driver 34, for example, the four solenoid valves 8 are closed. It is made to drive in order.

The functional block diagram for such fuel injection control (magnetic valve drive time control) is as shown in FIG. 1A. In other words, in view of the ECU 23 in software, the ECU 23 firstly has a basic drive time determining means 35 for determining a basic drive time T _B for the solenoid valve 8, which determines the basic drive time. The means 35 intakes per engine revolution with the intake air amount Q information from the air flow detector 11, the intake air amount Q information from the crank angle detector 11 and the engine speed Ne information from the crank angle detector 21. The air quantity Q / Ne information is obtained and the basic driving time T _B is determined based on this information. In addition, the air-fuel ratio up correction means 36 for performing the air-fuel ratio up correction according to the engine speed and the engine load (the Q / Ne information has engine load information) and the correction coefficient K _AF at the time of O ₂ sensor feedback are set and corrected. O ₂ detector feedback correction means (37) is provided and switching means (38), (39) for switching in conjunction with the air-fuel ratio up correction means (36) and O ₂ detector feedback correction means (37) Alternatively selected.

Furthermore, the cooling water temperature correction means 40 which sets the correction coefficient K _WT according to the engine cooling water temperature, the intake air temperature correction coefficient 41 which sets the correction coefficient K _AT according to intake temperature, and the correction coefficient K _AC according to atmospheric pressure are set. Dead weight correction means for setting the atmospheric pressure correction means 42, the acceleration weight correction coefficient 43 for setting the correction coefficient T _D for the acceleration weight, and the dead time (valid time) T _D for correcting the driving time according to the battery voltage. The means 44 is set, and at the time of O ₂ feedback correction, the driving time T _INJ of the solenoid valve 8 is finally set to T _B × K _WT × K _AT × K _AP × K _AF + T _D , and this time K _The solenoid valve 8 is driven by _INJ .

The control method for driving the solenoid valve is shown in the flow chart of FIG. 5, which is operated by the interruption of the crank pulse every 180 °, and the step (b1) shows what the fuel cut plug set is. It is determined that the fuel cut plug set does not need fuel injection and is returned. Otherwise, the crank angle 180 is based on the period data between the number of Kalman pulses and Kalman pulses generated between the previous crank pulse and the current crank pulse in step b2. Set the intake air volume Q _CR (Q / Ne) per °.

Then, in the next step (b3), the basic drive time T _B is set according to this Q _CR , and then in step (b4), the solenoid valve drive time T _{INJ is set} to T _B × K _WT × K _AT × K _AP × K _AF × T _It is calculated by operation from _D and triggers the injection timer on this T _INJ in step (b5). And, when triggered in this way, fuel is dispersed for the time T _INJ .

By the way, in the air-fuel ratio feedback control using the O ₂ detector, the output Vf from the upstream O ₂ detector 17 and the predetermined reference value Vfc (the intermediate value between the high level output and the low level output of the upstream O ₂ detector 17 of this reference value Vfc) When selected, it is possible to be rich when Vfc> Vf, as opposed to so-called rich and lean determination voltages. It is supposed to lean at Vf.

Therefore, the O ₂ sensor feedback correction means 37 compares the comparison means 46 with the reference value Vfc from the rich and low determination amount voltage setting means 45 that sets the reference value Vfc as shown in FIG. There is a correction coefficient determination means 47 for determining the air-fuel ratio correction coefficient K _AF according to the comparison result from the comparison means 46, which is different from the conventional one in the air-fuel ratio control device, and is an upstream O ₂ detector 17 and a downstream O _2. Based on the outputs Vf and Vr from the detector 18, a reference value changing means 48 capable of changing the reference value (rich-lean determination voltage) Vfc, for example, after a required traveling distance or a storage battery detachment history is provided.

Next, based on the amount of output Vf, Vr from the upstream O ₂ sensor 17 and the downstream O ₂ sensor 18 will be described with respect to the reason to change by modifying with an appropriate rich Lin determined voltage Vfc.

Now, if the output Vf of the upstream O ₂ detector 17 is taken as the horizontal axis, and the output Vr of the downstream O ₂ detector 18 is obtained as the relationship between the two outputs Vf and Vr in the vertical axis, the characteristics are as shown by the solid line in Fig. 8B. Was found. Comparing these characteristics with the NOx purification efficiency characteristics (see FIG. 8a solid line) and the Co and Hc purification efficiency characteristics (see dashed line FIG. 8a), the highest purification efficiency shown in FIG. 8a (where the theoretical air-fuel ratio) and It can be seen that the output value Vfc of the upstream O ₂ detector 17, which generates a drastic change in the characteristics shown in FIG. 8B, coincides.

In addition, the air-fuel ratio in which the Vr change is very large with respect to the change of the output Vf is high regardless of the unevenness or the secular variation of the characteristics of each product of the O ₂ detector, and the purification efficiency of the three components of Hc, Co, and NOx is high. It was found that the air-fuel ratio (theoretical air-fuel ratio).

Further, both output characteristics from the upstream O ₂ detector 17 and the downstream O ₂ detector 18 become as shown in FIG. 8B for the following reason. That is, if unburned components such as Co are present in the exhaust gas, the O ₂ detector output is shifted up overall. In this case, even if the air-fuel ratio is low, non-combusted gases such as Hc, Co, and H ₂ exist on the upstream side of the catalytic converter 9. The output of the upstream O ₂ detector 17 thus rises for this reason. On the other hand, since the non-combustion gas is purified by the catalyst switch 9 on the downstream side or inside of the catalytic converter 9, the output of the downstream O ₂ detector 18 does not rise. And since these relations are prominent in the vicinity of the theoretical air-fuel ratio, they become a characteristic as shown in FIG. 8B.

For this reason, the reference value changing means 48 has a characteristic calculating means 49 for calculating the characteristic between the outputs of both the upstream O ₂ detector 17 and the downstream O ₂ detector 18 as shown in FIG. 8B. The output value Vfc of the upstream O ₂ detector 17 obtained by the characteristic calculating means 49 is updated as a new rich-lean determination voltage Vfc. This update function is assumed to be included in the rich / lean determination voltage setting means 45.

The Vf-Vr characteristic and the reference value Vfc for rich and lean determination are geared to the BURAM 33.

Next, the main routine of this air-fuel-ratio control apparatus including the change of the said reference value, determination of a correction coefficient, etc. is demonstrated in detail using FIGS. 4A-4B. 4A to 4D show one flowchart, but since the flowchart is long, it is divided into four parts for the sake of convenience.

First, in this main flow, as shown in FIG. 4A, the system starts with the key switch (ignition switch) on and initializes the RAM 32 and the interface at step a1. Next, in step a2, it is determined whether the storage battery 24 is attached or detached. Normally, since the storage battery 24 is in the ON state, a "no" root is taken and the traveling distance data OD is input in step a3.

In step a4, this OD data is compared with the reference value rewriting distance ODX (this reference value rewriting distance ODX is not backed up by the battery), and not OD> ODX. In other words, when the reference value rewriting distance has not yet been reached, the operation state information is input in step a5, and it is determined whether the fuel cutoff region is in the next step a6.

If it is not the fuel cutoff area, the correction coefficients K _WT , K _AT , K _AP and K _Ac are set in step a8 after the fuel cut plug is reset in step a7, and the dead time T _D is set in step a9. do. These coefficients and the like are set by the cooling mercury correction means 40, the intake air temperature correction means 41, the atmospheric pressure correction means 42, the acceleration weight correction means 43, and the dead time correction means 44.

Next, in step a10, it is determined from the output voltage value whether the upstream O ₂ detector 17 is active.

Then, if the upstream O ₂ detector 17 is active, it is determined in the next step a12 whether the air-fuel ratio (A / F) feedback mode is. It is determined in the A / F feedback mode when the cooling water temperature is higher than the predetermined value in the required driving zone (A / F zone) determined by the engine load and the engine speed.

If it is in A / F feedback mode, it is determined in step a13 whether or not the O ₂ detector calibration check completed plug set is established. In general, since the jump is made in step a71, it is in the set state, so the "Yes" root is taken and the upstream O in step a14. Leanization is shown in ₂ (2).

After step a16-3, the feedback correction coefficient K _FB is obtained as 1 + P + 1 in step a17, and the value K _FB is put into the address K _AF in step a21. Initially, it starts from K _FB = 1 because proportional gain P = 0 and integration coefficient 1 = 0.

Next, the scanning counter is initially set in step a24. At this time, an initial value other than zero is selected. In addition, this scan counter is used at the time of a reference value change update mentioned later, and n Vf counters used simultaneously are also reset by this step a24.

In step a25, the cycle count SCOUNT (this is also used for the reference value change and update described later) is set to 0, and then the process returns to step a5.

If it returns to step a15 again, in this case, since the WOFB plug is reset in step a16-2, a "no" root is taken and it is determined whether plug L = 1 in step a16-4. In this case, since L = 1 at step a16-3, a "yes" route is taken and the process of step a17 is performed.

However, the integral time calculation routine for the integral coefficient 1 has the flowchart shown in Fig. 6B, in which the routine determines whether or not the WOFB plug is set at step d2 whenever the timer is interrupted (A / F feedback). Mode), it is determined whether or not the plug L = 1 in step d3. If L = 1, I _LR (LIN) is added to I in step d4. If not, newly subtracting I _RL (Lean Integration Coefficient) from I in Step d4 is performed as I. In this way, I _LR is added to every timer interrupt for L = 1 days, and I _RL is to be subtracted every timer interrupt while L = 1 is not (between L = 2). Therefore, while the I _LR is added, the feedback correction coefficient K _FB becomes large and richening is promoted, while while the I _RL is subtracted, the feedback correction coefficient K _FB becomes small and _leaning is made to increase.

In this case, I = _LR is added to each timer interrupt because L = 1, and K _FB of the feedback correction coefficient becomes large, so that the output Wf of the detector 17 and the rich / lean determination voltage Vfc are compared, and when Vfc> Vf, In (a15), it is determined whether or not the wise-out feedback plug (hereinafter referred to as WOFB plug) is a set. Immediately after entering the A / F feedback area, the WOFB plug is set, so take the "yes" route, set the proportional gain P to 0 in step (a16-1), and reset the WOFB plug in step (a16-2) Let plug L be 1 in a16-3).

Here, this plug L represents richening in (1), and richening is promoted.

Rich results in this way, Vfc When Vf is reached, a "no" root is taken in step a14, and it is determined whether the WOFB plug is set in step a18. In this case, since the WOFB plug is still in the reset state in the A / F feedback mode, a "no" root is taken in this step a18, and it is determined whether plug L is 2 in step a19-1. Immediately after the switching, L = 1, so that the rinsing proportional gain P _RL is subtracted from the proportional gain P in step (a19), and this is obtained as P, and the feedback correction coefficient R _FB is calculated as 1 + P + 1 in step a17 and this value K _FB is obtained in step a21 by address K _AF Put it in. As a result, the feedback correction coefficient K _FB is lowered by the lean proportional gain P _RL at the maximum state.

After that, the scan counter is initially set in step a24, and the cycle count SCOUNT is 0 in step a25, and then the process returns to step a5.

If the process returns to step a19-1 again through step a18, in this case, L = 2 in step a19-3, the process returns to step a17.

In this case, L = 2, so that every time the timer is interrupted, a "no" route is subtracted from step d2 of FIG. 6B, and the IRL is subtracted from step d4, and the feedback correction coefficient K _FB decreases, so that leaning is promoted.

As a result of leaning in this manner, when Vfc > Vf, a " yes " route is taken in step a14, and it is determined whether the WOFB plug is set in step a15. In this case, since the WOFB plug is still in the reset state in the A / F feedback mode, a "no" root is taken in this step a15, and it is determined whether plug L is 1 in step a16-4. . Since L = 2 immediately after switching, the feedback correction coefficient K _FB is obtained as 1 + P + 1 in step a17, and the value K _FB is put in the address K _AF in step a21. This _{raises the} feedback correction coefficient K _FB by the richening proportional gain P _LR from the minimum state.

Subsequently, by repeating the above process, the feedback correction coefficient K _FB is varied as shown in FIG. 7C, whereby the required air-fuel ratio control is executed in the A / F feedback mode.

FIG. 7A is an upstream O ₂ detector output waveform diagram, and FIG. 7B is a rich and lean determination waveform diagram.

In the case where Vfc ≤ Vf immediately after entering the A / F feedback area, the WOFB plug is set immediately after entering this case, so that the "yes" root is taken in step a18, and the proportional gain P in step a19-4. Is 0, the WOFB plug is reset in step (a19-5), and the plug L is 2 in step (a19-3). After the step a19-3, the feedback correction coefficient K _FB is obtained as 1 + P + 1 in step a17, and the value K _FB is put in the address K _AF in step a21. Therefore, also in this case, the proportional gain P = 0 and the integral coefficient 1 = 0 are also started from K _FB = 1.

Thus, the comparison means 46 and the correction coefficient determination means 47 in the O ₂ sensor feedback correction means 37 determine the correction coefficient K _AF based on the comparison between Vfc and Vf and the result of the comparison.

After the step a5, if the fuel cutoff region is reached in step a6, the fuel cutoff plug is set in step a27, the integration coefficient I is set to 0 in step a28, and the timer T again in step a29. _An initial value (e.g., equivalent to 10 seconds) is input to _KC , and in step a30, the mapped A / F correction coefficient K _AFU is set in accordance with the engine load or the engine speed, and this is put in the address K _AF in step a31, and the step (a31) After setting the WOFB plug in -2), the process returns to step a5 via steps a24 and 25a. Since the WOFB plug is set in step a31-2, the WOFB plug is set immediately after entering the A / F feedback mode.

In the case of "no" in steps a10 to a12, since A / F feedback control cannot be performed, steps (a5) pass through steps (a28 to a31), (a31-2), (a24) and (a25). Return to

In normal operation, the above routine is repeated to set the coefficients K _WT , K _AT , K _AP , K _AC , K _AF or time T _{D according} to the engine condition, and write these values to drive the solenoid valve shown in FIG. By operating the routine, the solenoid valve 8 is injecting a desired amount of fuel. In this way, the desired air-fuel ratio control is performed.

However, when the traveling distance OD becomes the reference value rewriting distance ODx, a "yes" routine is taken in step a4, and the O ₂ sensor correction check completed plug is reset in step a71.

After that, the processing of step a6 is performed via step a5. If the fuel is not a fuel cutoff region at step a6, the processing of steps a10 to a12 is performed via steps a10 to a12. If the steps a10 to a12 are YES, at step a13, it is determined whether or not the O ₂ sensor correction check completed plug set is set. In this case, since it is reset at step a71, a "no" root is taken and moved to steps a11, aa, and a33 shown in FIG.

In step a11, it is determined whether the downstream O ₂ detector 18 is active, and in steps a32 and a33, it is determined whether the engine speed Ne is 3000 rpm or less, or more than 1500 rpm. If yes, it is determined at step a34 whether the engine variation │dNe / dt│ is less than the setpoint DNx, and if it is small, at step aa35, a36 whether the intake air quantity Q is greater than the setpoint Qx, It is determined whether the intake air change │dQ / dt│ is smaller than the set value DQx, and if all are YES, it is determined at step a37 whether the change │dθ / dt│ of the throttle opening degree θ is smaller than the set value DTHx. . If YES in this step 37 as well, it is determined in step a39 whether the timer T _KC is zero.

In addition, the timer T _KC is processed every time the timer is interrupted by the timer subtraction routine shown in FIG. 6A. In other words, 1 is subtracted from the contents of T _KC in step c1 to be T _KC . That is, it is made to count down.

If this timer T _KC is not 0, the process returns to the process after step a14 shown in FIG. 4A.

Further, the step (a32 to a37) in both the "no" one cases, enter a value (such as those given in the step (a29)) The initial value of the timer T _KC in step (a38) and step shown in claim 4a Fig. (a14) It returns to the subsequent processing.

As a result, even if the mileage data OD becomes the reference value rewriting distance ODx, the two O ₂ detectors 17 and 18 are not active or the A / F feedback mode (the driving range in this mode is relatively stable. Is not set) or the engine speed Ne is (1500) Ne If the engine variation is large, the intake air amount is small, the intake air amount fluctuation, or the throttle opening degree is large, all of them do not move to the reference value rewriting process and return to the routine walk side during normal operation. have.

For example, even after all of the above conditions are satisfied, if a certain time (time equivalent to the initial value of the timer T _KC ) has not passed, the procedure does not shift to the reference value rewriting process, and the routine walk side in normal operation is performed. It is returned.

If all of the above conditions are satisfied and the time required after the condition is established (hereinafter, these conditions are referred to as reference value rewriting conditions), the step of FIG. 4C is made after setting the WOFB plug in step (a39-2). At (a49), it is determined whether or not the scan cycle counter is zero. Initially, since an initial value other than zero is set in step a24 of FIG. 4a, a "no" route is taken to determine whether or not the cycle number SCOUNT is zero in step a50. In this case, since it becomes 0 in step a24 shown in FIG. 4A, a "yes" route is taken, and in step a50, a decrement (DCR) process of subtracting 1 from the contents of the scan cycle counter is performed. The plug COND is set to 1 in the next step a52, and the state of the plug COND is determined in the step a53. In this case, since COND is 1, the cycle number SCOUNT is increased by one step at step a54.

After that, in step a55, the air-fuel ratio coefficient K _S is obtained as 1 + (1-SCOUNT / 128) x 0.05 (in this case, SCOUNT is 1, K _S ≒ 1.05), and in step a56, the coefficient K _{AF is set} to K _S. After obtaining and artificially making the rich side, in step a57, the output Vf of the upstream O ₂ detector 17 and the output Vr of the downstream O ₂ detector 18 are read out, and in step a58, the memory in which Vr is positioned as Vf ( RAM), and the number of data corresponding to Vf added in step a59 is increased by one.

In this case, the number of memory addresses is selected enough to create the Vf-Vr characteristic diagram shown in FIG. 8B, and the inverse of the number corresponds to the resolution, but the number of Vf counters is prepared by the number n corresponding to this memory address. If it is stored at the corresponding address, the count is increased by one.

After step a59, the process returns to step a5 of FIG. 4a and passes again through the "no" route in step a6, the "no" route in step a13, and the "yes" route in step a39. Returning to step a49 shown in Fig. 4c, the scan cycle counter is not yet zero, so a "no" route is taken to determine whether or not SCOUNT is zero at step a50. In this case, since SCOUNT is set to I in step a54, a "no" root is taken in step a50, and it is determined whether or not SCOUNT is (255) in step a60. In this case, since it is "no", it jumps to step a61 and determines in what state the plug COND is in step a53. In this case, since the state in which COND is 1 in step a52 has not been eliminated, SCOUNT is increased by 1 again in step a54. This sets the coefficient K _S by setting the part of SCOUNT / 128 to 2/128 in step (a55), and then obtain the coefficient L _AF to be slightly lean side, and then the upstream O ₂ detector (17) and the downstream O ₂ detector. Step (a5) of Fig. 4a after reading the outputs Vf and Vr of (18), adding Vr to the memory positioned at Vf, and increasing the number of data corresponding to the added Vf (steps a56 to a59). In the same manner as above, the process returns to step a49 of FIG. 4C again.

Thereafter, the above process is repeated until SCOUNT = 255. This is in turn operated from the rich side to the lean side (about 1.05 to 0.95 as the value of K _S ), and the theoretical air-fuel ratio is read by reading the respective outputs Vf and Vr of the upstream O ₂ detector 17 and the downstream O ₂ detector 18 at that time. The Vf-Vr characteristic when operating from the rich side to the lean side in the vicinity can be measured.

When SCOUNT is 255, the plug COND becomes 0 (step a61) since the switching to the "yes" route is performed at the step a60.

Therefore, the processing in step a62 is performed next in step a53. That is, the cycle number SCOUNT is reduced by one step.

Then, in step a55, the air-fuel ratio coefficient K _{S is obtained} from 1 + (1-SCOUNT / 128) x 0.05 (in this case, SCOUNT is 254, so K _S ≒ 0.95), and in step a56, the coefficient K _AF is calculated as K _S. Then, at step a57, the output Vf of the upstream O ₂ detector 17 and the output Vr of the downstream O ₂ detector 18 are read and at step a58, Vr is added to the memory (RAM) positioned by Vf. The number of data corresponding to Vf added in step a59 is increased by one. In this case, since it is the 2nd time, setting the count of the corresponding counter to 2 is performed.

Next to step a59, the process returns to step a5 of FIG. 4a, passing the "no" root at step a6, the "no" root at step a13, and the "yes" root at step a39. Returning to step a49 shown in FIG. 4C again, since the scan cycle counter is not yet zero, a "no" route is taken and it is determined whether or not SCOUNT is 0 in step a50. In this case, since SCOUNT is 254 at step a62, a "no" root is taken at step a50, and it is determined whether SCOUNT is 255 at step a60. In the present case, since it is "no", it jumps to step a61 and determines in what state the plug COND is in step a53. In this case, since the state in which COND is 0 in step a61 is not solved, SCOUNT is reduced by 1 again in step a62. In step a55, the coefficient K _S is set by setting the portion of SCOUNT / 128 to 253/128, after which the coefficient K _AF is obtained, and then the upstream O ₂ detector 17 and the downstream O ₂ detector 18 are obtained. After reading each output Vf, Vr of V, add Vr to the memory defined for Vf, increase the number of data corresponding to the fresh Vf (steps a56 to a59), and then return to step a5 of FIG. In the same manner, the process returns to step a49 of FIG. 4C again.

Thereafter, the above process is repeated until SCOUNT = 0. In this way, the lean side is operated from the lean side to the rich side (about 0.95 to 1.05 in the value of K _S ), and the respective outputs Vf and Vr of the upstream O ₂ detector 17 and the downstream O ₂ detector 18 are read out, and then near the theoretical air-fuel ratio. The Vf-Vr characteristic of the 2nd time when operating from the lean side to the rich side can be measured. This results in a one-reciprocal measurement of the theoretical air-fuel ratio (Vf-Vr characteristics in the range of 1.05 and 0.95 in the value of K _S ).

When SCOUNT becomes 0, the switch is switched to " yes " root at the step a50, so the plug COND is set to 1 after the scan cycle counter is reduced by one (step a52).

Therefore, after that, the operation of the lean side or the reverse is reciprocated again from the rich side, and the Vf-Vr characteristics of the third and fourth times are measured.

Then, if the above Vf-Vr characteristic planning operation is performed three or four round trips (this round trip number is based on the initial value set in the scan cycle counter), the value of the scan cycle counter becomes zero in step a51, and then the step (after When returning to a49), the "yes" root is taken and the process of step a63 shown in FIG. 4d is performed. That is, in this step a63, the average value of Vr with respect to (Vf) i which has been measured so far r [(Vf)] is calculated. The count number of the Vf counter is used when calculating this average value.

In this way, when the average value of Vr is obtained, the Vr-Vf curve is smoothed using interpolation or the like appropriate in step a64. The characteristic (refer FIG. 8c) obtained at this time is set as the Vf-Vr characteristic shown in FIG. 8b.

In step a65, the Vf range where dVr / dVf> K is obtained, i.e., the portion where Vr rises rapidly, and in step a66, the new value Vfc is used as the reference value Vfc for the rich-lean determination of the Vf range. It is stored in (33). This completes the replacement of the reference value Vfc, that is, the update of the reference value Vfc. Thereafter, at step a67, the O ₂ sensor correction check completion plug is set, at step a68, the mileage data OD is input, and at step a69, the next reference value rewriting distance ODX is, for example, ODX + 800 (mile). do.

After that, the process returns to step a5 of FIG. 4a, and if it is not the fuel cutoff region, a "no" root is taken at step a6. After the processing of steps a7 to a9, for example, all of steps a10 to a12 are " YES ", it is determined in step a13 whether or not the O ₂ detector calibrated check completed plug set is set. In step a67 of FIG. 4D, since the plug is in the set state, the aforementioned normal operation defined after step a14 is performed. Perform a routine walk of the city.

In this case, air-fuel ratio control is performed based on the rich-lean determination reference value Vfc updated as described above.

In this manner, the reference value Vfc for the rich and lean determination to be compared with the output Vf of the upstream O ₂ detector 17 is changed and updated based on the respective outputs Vf and Vr of the upstream detector 17 and the downstream O ₂ detector 18. As a result, the control accuracy does not change with the characteristic unevenness of each product of the O ₂ detector and the secular variation of the characteristics, and the exhaust gas purification efficiency by the catalytic converter 9 can also be kept high, thereby obtaining higher control reliability. Lose.

In addition, even when an EGR is not performed or even an EGR is performed, even in the case of a scale, a good exhaust gas level is obtained, so that the EGR system can be simplified, and the exhaust gas does not sacrifice power performance or operation function.

In addition, the rich-lean determination voltage Vfc is stored in the BURAM 33, and the memory value does not disappear only by turning off the ignition switch 26. However, when the battery 24 is removed, the stored contents are also turned off. ° step (a2) the O ₂ sensor throne check completion plug from the value of a representative Vf in step (a70) if the history of the battery removal (e.g. 0.6 volts significant amount) temporarily enter a as an initial value, and thereafter, step (a7) Lee in The setting is done.

When the O ₂ detector calibration check completed plug is reset in this manner, a "no" root is taken in step a13 to satisfy the reference value rewrite condition, and then the reference value Vfc for rich-lean determination is changed. In this case, the description is the same as in the case of the battery reference value rewriting, and thus the detailed description is omitted.

Further, in this embodiment, the air-fuel ratio feedback system is made from the closed loop to the open loop, and then the air-fuel ratio is changed in the vicinity of the theoretical air-fuel ratio by changing K _S stepwise, and Vf and Vr are set for a predetermined time in each air-fuel ratio. Each measurement was made and averaged, and the graph of FIG. 8c was obtained. Vf-Vr characteristics may be different from those in case of air-fuel ratio fluctuations as in actual feedback. The air-fuel ratio may be changed every 1/128 to 126/128 to 131/128 times the value of K _S while giving the air-fuel ratio fluctuation range 5% in terms of fuel.

9 to 12 show an air-fuel ratio control apparatus for an internal combustion engine as a second embodiment of the present invention. FIG. 9 is a main control block diagram, and FIGS. 10A to 10E are flowcharts for explaining the main routine, and FIG. FIG. 12 is a graph for explaining the feedback correction coefficient, and FIG. 12 is a graph for explaining the response time of the O ₂ detector. In FIGS. 9 to 12, the same reference numerals as those of FIGS. 1 to 8 show almost the same parts or processes. .

Now, the air-fuel ratio control device of the internal combustion engine as the second embodiment is in addition to that of the first embodiment described above, the rich-to-lean response time τ _RL and the lean-rich response time τ _LR of the upstream O ₂ detector 17. And either of the delay times DLYRL, DLYRL, the proportional gain DLYRL, DLYLR of the air-fuel ratio feedback, and the integral gains of the air-fuel ratio feedback P _RL , P _LR according to these response times τ _RL , τ _LR . It is.

Where the response time τ _RL is the time from the rich to the lean (A / F) ^c of the upstream O ₂ detector 17, and then the upstream O ₂ detector output Vf reaches the reference value Vfc, and the response time τ _LR is the lag-to-rich delay time of the upstream O ₂ detector 17 and the intake air-fuel ratio crosses the lean to reach furnace (A / F) c, and then the upstream O ₂ detector output Vf reaches the reference value Vfc. Refers to the time until [see FIG. 12A, FIG. 12B].

However, in the second embodiment, the air-fuel ratio feedback control using the O ₂ detector also requires the output Vf from the upstream O ₂ detector 17 and the reference value Vfc (this reference value Vfc is equal to the high level output of the upstream O ₂ detector 17). Is selected, and functions as a so-called rich and lean determination voltage). Rich when Vf and lean when Vfc> Vf. Therefore, as shown in FIG. 9, the O ₂ detector feedback correction means 37 outputs the output Vf and the rich V from the rich-lean determination voltage correction means 45 and the upstream O ₂ detector 17 which set the reference value Vfc. Comparison means 46 for comparing the reference value Vfc from the lean determination voltage setting means 45, and correction coefficient determination means 47 'for determining the air-fuel ratio correction coefficient K _AF in accordance with the comparison result from the comparison means 46. Although this air-fuel ratio control device is different from the conventional method, the reference value (rich lean determination voltage) Vfc is required, for example, based on the outputs Vf and Vr from the upstream O ₂ detector 17 and the downstream O ₂ detector 18. It has the reference changing means 48 which can change for every distance.

Then, the correction factor determining means (47 '), the response time τ _RL, to obtain a τ _LR these response time τ _RL, thus the response delay time on the τ _LR DLYRL, DLYLR, proportional gain P _RL, P _LR and integral gain I _RL and it has a means for correcting any one of the I _LR.

Further, the response delay time DLYLA, DLYLR, proportional gain P _RL , P _LR , integral gain I corrected according to the Vf-Vr characteristic, Vf-Ko characteristic, the reference value Vfc for the rich-lean determination, or the response time τ _RL , τ _LR . _RL and I _LR are stored in the BURAM 33.

Next, the main routine of this air-fuel ratio control apparatus including the change of the reference value, the determination of the correction coefficient, and the like will be described with reference to FIGS. 10A to 10E. 10A to 10E show one flow chart, but since the flow chart is long, it is divided into five diagrams in appropriate parts for convenience.

First, in this main flowchart, as shown in FIG. 10A, the system starts at the key switch (ignition switch) on and initially initializes the RAM 32 or the interface at step a1. Next, in step a2, it is determined whether the storage battery 24 is attached or detached. Normally, since the battery 24 sms is as good, a "no" route is taken and the traveling distance data OD is input in step a3.

In step a4, the OD data is compared with the reference value rewriting distance ODX (this reference value rewriting distance ODX is backed up by the battery), and is not OD> ODX. That is, when the reference value rewriting distance has not yet reached, the driving state information is input in step a6, and it is determined whether the fuel cutoff region is in the next step a6. And, if not the fuel cut region yen set the step (a7) a correction coefficient in step (a8) of the fuel shut-off plug Li seobuteo set hagwan in K _WT, K _AT, K _AP, K _Ac and dead time in the step (a9) Set T _D. These coefficients and the like are set by the cooling water temperature correction means 40, the intake air temperature correction means 41, the atmospheric pressure correction means 42, the acceleration weight correction means 43, and the dead time correction means 44.

Next, at step a10, it is determined from the output voltage pole whether the upstream O ₂ detector 17 is in an active state. If the upstream O ₂ detector 17 is active, it is determined in the next step a12 whether it is the air-fuel ratio (A / F) feedback mode.

If it is the A / F feedback mode, at step a13, it is determined whether the feedback characteristic value (FB characteristic value) calculation complete plug set is established, and since it is normally in the set state, a "yes" route is taken and upstream O ₂ at step a14. The output Vf of the detector 17 and the rich-lean determination voltage Vfc are compared, and when Vfc > Vf, it is determined at step a15 whether the WOFB plug is set. Immediately after entering the A / F feedback area, the WOFB plug is set, so the "yes" root is taken, the proportional gain P is set to 0 at step a16-1, and the WOFB plug is reset at step a16-2. Let plug L be 1 in a16-3).

After the step a16-3, the feedback correction coefficient K _FB is obtained as 1 + P + 1 in step a17, and the value K _FB is put in the address K _AF in step a21. Initially, it starts from K _FB = 1 because the proportional gain P = 0 and integral coefficient 1 = 0.

Thereafter, at step (a22), it is a reset at first to determine whether or not it is a Kc count start plug set, so it jumps to step (a23-2) and determines whether it is an O ₂ detector corrected check completed plug set. Usually, since it is in the set state, a "yes" route is taken and the process returns to step a5.

If it returns to step a15 again, in this case it is step a16-2 and since the WOFB plug has been reset, a "no" root is taken and it is determined whether plug L = 1 in step a16-4. In this case, since L = 1 at step a16-3, a "yes" route is taken and the process of step a17 is performed.

Incidentally, the integration time calculation lupine for the integral coefficient 1 is the same as the flowchart of FIG. 6B in the first embodiment described above.

In this case, since L = 1, I _LB is added to each timer interrupt, and the feedback correction coefficient K _FB becomes large, so that richening is promoted.

Rich results in this way, Vfc When Vf is reached, a " No " loop is taken in step a14, and it is determined whether or not the WOFB plug is set in step a18. In this case, if it is still in the A / F feedback mode, since the WOFB plug is still in the reset state, a "no" root is taken in this step 18, and it is determined in step a19-1 whether the plug L is 2 or not. do. Immediately after the changeover, L = 1, so take "No" root in step (a19-1) and Vfc in step (a19-1) It is determined whether the delay time DLYLR has elapsed after Vf, and while the delay time DLYLR has not elapsed, a "no" route is taken and the processing of step (a17) is performed. In step (a19-2), rinse proportional gain P _RL is subtracted from proportional gain P, and this is subtracted from rinse proportional gain P _RL from P, and this is P, and L = 2 in step (a19-3). In a07), the feedback correction coefficient K _FB is obtained as 1 + P + 1, and at step a21, the value K _FB is put in the address K _AF . As a result, the feedback correction coefficient K _FB is lowered by the lean proportional gain P _RL from the maximum state.

After that, the process returns to step a5 in the same manner as described above.

Then, if the process returns to step a19-1 again through step a18, in this case L = 2 in step a19-3, a "yes" route is taken and the process of step a17 is performed.

In this case, since L = 2, the "no" root is subtracted from step d2 of FIG. 6B, and I _RL is subtracted from step d4 for each timer insertion, and leaning is promoted by decreasing the feedback correction coefficient K _FB .

As a result of leaning in this way, when Vfc > Vf, if " Yes " is taken in step a14, it is determined whether or not the WOFB plug is set in step (a-15). In this case, since the WOFB plug is still in the reset state in the A / F feedback mode, a "no" root is taken in this step a15, and it is determined whether the plug L is 1 in step a16-4. . Immediately after the switching, L = 2, so it is determined whether or not the plug L is 1 in step a16-4. Immediately after the switching, L = 2 takes the “no” route in step (a16-4), takes the “no” route in step (a16-4 '), and the delay time after Vfc > Vf in step (a16-4'). It is determined whether or not DLYRL has elapsed. While the delay time DLYRL has not elapsed, a process of step (a17) is performed by taking a "no" route, but if the delay time DLYRL has elapsed, a "yes" route is taken and the step (a16-3) is performed. Add the proportional gain P _LR to P, set it as P, and set L = 1 in step (a16-3), then obtain the feedback correction coefficient K _FB as 1 + P + 1 in step (a17), and address this value K _FB in step (a21). Put it in K _AF . This raises the feedback correction coefficient K _FB by the lean scaling proportional gain P _LR from the minimum state. Thereafter, by repeating the above process, the feedback correction coefficient K _FB fluctuates as shown in FIG. 11C, and thus the required air-fuel ratio control is executed in the A / F feedback mode.

Moreover, the 11a to turn an output waveform upstream O ₂ sensor also, the 11b turns rich-lean judgment waveforms and delay time DLYLR is just as illustrated in 11a also O ₂ sensor output over the rich-lean judgment voltage Vfc from below or It is the time equivalent to the time from when it traverses from the top to the rich lean determination as shown in FIG. 11B.

Also, Vfc immediately after entering the A / F feedback area In the case of Vf, since the WOF plug is set immediately after entering this case, the "Yes" root is taken in step (a18), the proportional gain P is set to 0 in step (a19-4), and the WOFB plug in step (a19-5). Is reset, the WOFB plug is reset in step (a19-5), and the plug L is set to (2) in step (a19-3). After the step a19-3, the feedback correction coefficient K _FB is obtained as 1 + P + 1 in step a17, and the value K _FA is put into the address K _AF in step a21. Therefore, also in this case, the first time is proportional gain P = 0 and the integration coefficient I = 0 and start again from K _FB = 1.

Thus, the comparison means 46 and the correction coefficient determination means 47 'in the O ₂ sensor feedback correction means 37 determine the correction coefficient K _AF based on the comparison between Vfc and Vf or on the basis of this comparison result.

In the present embodiment, the delay times DLYRL, DLYLR, proportional gain P _RL , P _LR , integral gain I _RL , I _LR are variable as described below.

After the step a5, when the fuel cutoff region is reached in step a6, the fuel cutoff plug is set in step a27, the integration coefficient I is set to 0 in step a28, and then the timer is reached in step a29. An initial value (e.g., equivalent to 10 seconds) is input to T _RC to set the mapped A / F correction coefficient K _AFM according to the engine load or the engine speed in step a30, which is input to the address K _AF in step a31. After setting the WOFB plug in step a31-2, the process returns to step a5 via step a23-2.

In the case of "no" in steps a10 to a12, since A / F feedback control cannot be performed, the process returns to step a5 through steps a28 to a31, a31-2, and a28-2. Goes.

During normal operation, the above-described routines are repeated to set the coefficients K _WT , K _AT , K _AP , K _AC , K _AF or time T _D in accordance with the state of the engine. Similarly, the solenoid valve 8 injects a desired amount of fuel by operating the solenoid valve drive routine shown in FIG. In this way, the desired air-fuel ratio control is performed.

However, when the driving distance OD becomes the ODX of the reference value rewriting distance, the "Yes" root is taken in step (a4), and the O ₂ sensor correction check completion plug is reset in step (a71), and the FB characteristic value is calculated in step (a71-2). The complete plug set is reset.

After that, the processing of step a6 is performed through step a5. If the fuel blocking area is not included in step a6, the processing of steps a10 to a12 is performed via steps a7 to a9. In the case of YES in steps a10 to a12 as well, it is determined in step a13 'whether the FB characteristic value calculation complete plug set is made. In this case, since it is reset in step a71-2, a "no" root is taken and moved to steps a11, aa, and a33 shown in FIG.

In step a11 it is determined whether the downstream O ₂ detector 18 is active, and in steps a32 and a33 it is determined whether the engine speed "no" is less than or equal to 3000 rpm or more than 1500 rpm. If both are "Yes", it is determined at step a34 whether the engine variation | dNe / dt | is smaller than the set value DN _r , and if it is small, the intake air quantity Q is set value Q at steps (a35) and (a36). _It is determined whether more than _x or the intake air change │dQ / dt│ is less than the set value DQ _x , and if both are “Yes”, the change of throttle opening degree θ dθ / dt│ is set value DTH _x in step (a37). It is determined whether it is smaller. If YES in this step a37 also, it is determined in step a39 whether the timer T _RC is zero.

In addition, the timer T _{KC is} also processed every timer interrupt by the timer subtraction routine shown in FIG. 6A as in the first embodiment.

And the timer T _KC is, if not zero, reset the K _C count start plug in the step (a40) and the coefficient in the step (a41) K _C (the coefficient K _C is A / F when the feedback value is theoretical air-fuel ratio And set the initial value other than the cycle count 0 in step (a41-2) with 1 as the same as the first embodiment described above, and then proceed to the process after step a14 shown in FIG. 10a. Go back.

Then, after passing through the steps a14 to a21 and passing through the "no" root at step a22 and coming to step a23-2, at this time, since the O ₂ sensor calibration check completion plug is reset at step a71, "no". The route is taken and the scan counter is initially set in step a24. At this time, the initial value is selected a suitable number other than zero.

This scan counter is used for the reference value change update similarly to the first embodiment, and n Vf counters used at the same time are also reset to this step a24.

Further, at step a25, the cycle number SCOUNT (which is also used for the reference value change update) is set to 0, and at step a26, the K _c count end plug is reset, and the flow returns to step a25.

The reset of the Vf counter may be performed in step a41-2. In addition, if all are "no" in steps a32 to a37, the initial value (the same value as given in step a29) is input to timer T _Kc in step a38, and the count Kc in step a40. The start plug is set, the coefficient Kc is set to 1 in step a41, and then the initial value is set in the cycle counter in step a41-2, and the process returns to the process after step a14 shown in Fig. 10A.

In this way, even if the mileage data OD becomes the reference value rewriting distance ODX, both O ₂ detectors 17 and 18 are not active or the A / F feedback mode (the driving range in this mode is relatively stable). Engine speed "Ne" is 1500 Ne When it is not 3000, or when the engine variation is large, the intake air amount is small, or the intake air amount variation or the throttle opening degree variation is large, all of them are not transferred to the standard value rewriting process, but are returned to the routine walk side in normal operation.

For example, when the above conditions are completely satisfied and the required time has elapsed after these conditions are established (these conditions are referred to as reference value rewriting conditions as in the above-described first embodiment), O ₂ at step a42 'in FIG. 10b. It is determined whether the detector calibration check is complete. In this case, since this plug is reset in step a71, a "no" route is taken and it is determined in step a42 whether it is a Kc count end plug set.

First, since the count end plug is reset at step a26 (see FIG. 10A), a "no" route is taken, and it is determined whether the Kc count start plug is set at the next step a43. Initially, the K _c count start plug is in the reset state and takes a "no" route to determine whether the coefficient K _FB is the maximum value K _FB (EXT). If the maximum value K _FB (EXT) is reached, the K _c count start plug is set in step a45 to perform the process after step a14 in FIG. 10a. In this case, the step (a15 to a21) post-treatment, the step (a22) in K _c count start plug set whether to determine whether it is K _c count start plug at the 10b-degree step (a45) is set In the step (a22) The value (median value) K _C to be the theoretical air-fuel ratio when A / F feedback is taken in the next step (a23) with the "Yes" root, and then KK _C + (1-K) (K _FB -1). The process returns to step a23-2, (a24 to a26), and (a5).

In the case of "no" in step a44, that is, while the coefficient K _FB is not reaching the maximum value, step a45 is jumped to step a14. The process returns to step (a23-2), (a24 to a26), and (a5) to jump, so that the update correction of the median value K _C is not performed.

Thereafter, similarly, when it comes to step a43 again, since the K _c count start plug is set in step a45, the "Yes" route is taken and the coefficient K _FB detects the first maximum value, and then the fourth maximum value K _FB ( EXT) is determined in step a46.

While the fourth maximal value K _FB (EXT) is not reached, take the "No" route in step a46, go through the steps a14 to a21, take the "Yes" route in step a22, and update the median K _C. The processing after steps a23-2 and (a24) is corrected.

In the meantime, if the reference value rewriting condition is not satisfied, the median value K _C is temporarily set to one.

When the fourth maximum value K _FB (EXT) is obtained with respect to the coefficient K _FB , the Kc count start plug is reset in step a47 and the K _c count end plug is set in step a48 (a42). Return to At this time, the median value K _C stores the average value for four times at a required address. In this way, the process of calculating the average value of the median value K _C is also referred to as pre-replacement.

In this way, when the preprocessing of the reference value is over, the "Yes" root is taken in step a42, and the scan cycle counter is 0 in step a49 of FIG. 10c after setting the WOFB plug in step a42-2. Is determined. Initially, a non-zero initial value is set in step a24 of FIG. 10a, and a "no" route is taken to determine whether or not the cycle number SCOUNT is zero in step a50. In this case, since it becomes 0 in step a25 shown in FIG. In the next step a52, the plug is COND as 1, and the plug COND is determined in step a53. In this case, since COND is 1, the number of cycles COUNT is increased by one step at step a54.

Then, in step (a55), the air-fuel ratio coefficient K _{S is obtained} from 1 + (1-SCOUNT / 128) x 0.05 (in this case, SCOUNT is 1, L _S ≒ 1.05), and in step (a56 '), the coefficient K _{C is obtained} by K. Obtain from _S × K _C (in this case K _C is 1 or almost 1 value), or leave K _AF = K _O at step (a56 ") to artificially reach the rich side, then at step (a57) the upstream O ₂ detector ( The output Vf of 17) and the output Vr of the downstream O ₂ detector 18 are read out, and at step a58, Vr is added to the memory (RAM) positioned at Vf and the data corresponding to Vf now added at step a59. The number increases by 1. In this case, the memory address number is selected to be sufficient to produce the Vf-Vr characteristic diagram shown in FIG. 8B described in the first embodiment, and the inverse of the number corresponds to the resolution. If the number of Vf counters is prepared by the number n corresponding to the memory address number and stored in the corresponding address, the count number is changed. It is similar to the first embodiment described above that the number is increased by one.

Further, (see the process of the step (a58)) memory for Vr and Vf, but may use a common counter for (see the process in Step (a58-2)) memory for K _O may be used for only counter Vf, respectively.

Next to step a59, the process returns to step a5 of FIG. 10a, passing the "no" route at step a6, the "no" route at step a13, and the "yes" route at step a42. Returning to step a49 shown in FIG. 10C again, since the scan cycle counter is not yet zero, a "no" route is taken to determine whether or not SCOUNT is zero at step a50. In this case, since SCOUNT is 1 at step a54, a "no" route is taken at step a50, and it is determined whether or not SCOUNT is (255) at step a60. In this case, since it is "no", it jumps to step a61 and determines in what state the plug COND is in step a53. In this case, since the state in which COND is 1 in step a52 has not been eliminated, SCOUNT is increased by 1 again in step a54. This sets the coefficient K _S with the part of SCOUNT / 128 at 2/128 in step (a55) .After that, the coefficients K _O and K _AF are obtained and set slightly to the lean side, and then the upstream O ₂ detector (17) and the downstream O ₂ from then reads the respective output Vf, Vr of the detector 18, and adds the Vr and K _O at a position determined memory to Vf, increasing the data number corresponding to a sum Vf (step a56 to a59), also the It returns to step a5 of FIG. 10a, and it returns to step a49 of FIG. 10c again similarly to the above.

After that, the above process is repeated until SCOUNT = 255. By operating in this way from the rich side to the lean side (about 1.05 to 0.95 as the value of K _S ) and reading the outputs Vf and Vr of the upstream O ₂ detector 17 and the downstream O ₂ detector 18 at that time, in the air-fuel ratio it can be measured near the Vf -Vr and Vf characteristics -K _O characteristics when it went toward the lean operation in the rich side.

When SCOUNT is 255, the plug COND becomes 0 because it is cut to the "yes" root at the position of step a60 (step a61).

Therefore, in step a53, the processing of step a62 is performed next. That is, the cycle number SCOUNT is reduced by one step.

Then, in step a55, the air-fuel ratio coefficient K _{S is obtained} from 1 + (1-SCOUNT / 128) x 0.05 (in this case, SCOUNT is 254, so K _S ≒ 0.95), and in step a56 ', the coefficient K _O is K _S × Obtained from K _C and also set K _AF = K _O at step a56 ", and then at step a57 the output Vf of the upstream O ₂ detector 17 and the output Vr of the downstream O ₂ detector 18 are read and In steps a58 and a58-2, Vr and K _O are respectively added to the memory (RAM) positioned at Vf and the number of data corresponding to Vf now added in step a59 is increased by 1. In this case, Since is the second time, setting the count of the corresponding counter to 2 is performed.

After this step a59, the process returns to step a5 of FIG. 10a, and passes again a "no" root at step a13, a "no" root at step a13, and a "yes" root at step a42. Returning to step a49 shown in FIG. 10C, since the scan cycle counter is not yet zero, a "no" root is taken, and it is determined whether or not SOUNT is zero in step a50. In this case, since SCOUNT is 254 at step a62, a "no" root is taken at step a50, and it is determined whether SCOUNT is 055 at step a60. In the present case, since it is "no", it jumps to step a61 and determines in what state the plug COND is in step a53. In this case, since the state in which COND is 0 in step a61 is not solved, SCOUNT is reduced by 1 again in step a62. This sets the coefficient K _S with the portion of SCOUNT / 128 at 253/128 in step (a55), after which the coefficients K _O , K _AF are obtained and then the upstream O ₂ detector (17) and the downstream O ₂ detector (18). After reading the thrusts Vf and Vr of V and adding Vr and K _O to the memory positioned at Vf, and increasing the number of data corresponding to the added Vf (steps a56 to a59), step aa of FIG. 10a In the same manner as above, the process returns to step a49 of FIG. 10C again.

Thereafter, the above process is repeated until SCOUNT = 0. In the lean side is gradually thereto the rich side as the value of K _S 0.95 to 1.05 or so) to the operation by the time the upstream O ₂ sensor 17, and each output of the upstream O ₂ sensor (18) Vf, the stoichiometric air-fuel ratio hameuroseo reading the Vr a second Vf -Vr and Vf characteristics -K _O characteristics of when he's operation toward the lean side from the support in the vicinity can be measured. This is in the Vf-Vf and Vr Characteristics _K-O-running characteristics to the stoichiometric air-fuel ratio (near the range of 1.05 and 0.95 degree as the value of K _S) to be measured by one round trip time.

When SCOUNT becomes 0, the cycle is switched to " yes " at step a50. Then, the scan cycle counter is reduced by one, and the plug COND is set to one (step a52).

Therefore, after that, the lean side or the reverse operation is reciprocated again from the rich side, and the Vf-Vr characteristics and the Vf-K _O characteristics of the third and fourth times can be measured.

Then, if the above Vf-Vr characteristic and Vf-K _O characteristic measurement operations are performed for several round trips (this round trip frequency is based on the initial value set in the scan cycle counter), the value of the scan cycle counter is zero at step a51. In this case, when returning to step a49 again, a "yes" route is taken and the process of step a63 shown in FIG. 10d is performed. That is, in this step a63, the average value Vr [(Vf) i] of Vr with respect to the previously measured (Vf) i is calculated. In addition, the count number of a Vf counter is used at the time of calculating this average value.

In this manner, when the average value of Vr is obtained, the Vr-Vf curve is smoothed by using interpolation or the like appropriate in step a24. The characteristic obtained at this time (refer also to 8c of 1st Example) is set as Vf-Vr characteristic shown in FIG. 8b in 1st Example similarly.

In step a65, the position where dvr / dvf > K becomes, and the Vf range, that is, Vr rises rapidly, is obtained. In step a66, the new value Vfc is used as the reference value Vf for the rich-lean determination as the reference value Vf. It is stored in the BURAM 33. This completes the replacement of the reference value Vfc, that is, the update of the reference value Vfc.

In addition, the setting, and set the O ₂ sensor calibration check completion plug in the step (a67) by the K _O corresponding to the Vfc in step (a55-2) by Koc.

After that, the process returns to step a5 of FIG. 10a, and if it is not the fuel cutoff region, a "no" root is taken at step a6. After the processing of steps a7 to a9, all of the steps a10 to a12 are temporarily If YES, it is determined in step a13 'whether the FB characteristic value calculation complete plug set is made. In step a71-2, since this plug is in the reset state, it is still "No" root in this step a13'. If the reference value rewriting condition is satisfied, it is determined whether or not the step (a42 ') O ₂ sensor correction check completion plug is set. In this case, after the reference value update is completed in step a67 of FIG. 10a, since it is already set, a "yes" route is taken, and it is determined whether or not the cycle counter is zero in step a72 of FIG. 10e. In this case, since a non-zero initial value is set in step a41-2 of FIG. 10B, a "no" root is taken and it is determined in step a73 whether it is the rich mode or the lean mode. If it is the rich mode, the coefficient K _AF = K _OC x 1.1 is set at step a74, and it is determined at step a75 whether the upstream O ₂ detector 17 is inverted from lean to rich.

Thereafter, the response time τ _LR equivalent value DTLR from the lean to the rich of the upstream O ₂ detector 17 is measured at step a76, and the content of the cycle counter is reduced by I at step a77 to step aa of FIG. Return to

Here, the measurement of the DYLR is performed by telling the injection command to the solenoid valve 8 and then measuring the time from the lean to the rich of the upstream O ₂ detector 17. For example, the DYLR measurement counter is always reset until the injection command is issued. When the injection command is issued, the up count or down count is outputted. Then, when the output of the upstream O ₂ detector 17 is reversed from lean to rich, the count is stopped. The value at that time is latched as DYLR.

If it is the lean mode in step a73, the coefficient K _AF = K _OC is set in step a78, and in step a79, it is determined whether the upstream O ₂ detector 17 is inverted from rich to lean.

Thereafter, the response time τ _RL equivalent value DYLR from the rich to lean of the upstream O ₂ detector 17 is measured at step a80, and the content of the cycle counter is reduced by 1 at step a77 to step a5 of FIG. Return to

Here, the DYLR is measured by issuing an injection command to the measurement diagram solenoid valve 8 and then measuring the time from the upstream O ₂ detector 17 to inversion from rich to lean. In this case, for example, the DYLR measurement counter is always reset until the injection command is issued, and when the injection command is issued, an up count or a down count is generated, and then the output of the upstream O ₂ detector 17 is inverted from rich to lean. The count stops and the current value is latched as DYLR.

In this way, when the cycle counter becomes 0 repeatedly by repeatedly measuring the DYLR and the DYLR, a "yes" route is taken in step a72, and the average value of the DYLR and DYLR is calculated in step a81.

In this way, the response times τ _RL and τ _LR of the upstream O ₂ detector 17 are obtained. Such response times τ _RL and τ _LR are obtained from a load obtained by obtaining the Vf-Vr characteristic shown in FIG. 8B as described above. 12a is also obtained by giving a periodic air-fuel ratio mode as shown in (b) with the air-fuel ratio feedback open looped. Here, the center value A / F of the air-fuel ratio fluctuation mode shown in FIG. 12A corresponds to the center value Kc from which Vfc is obtained.

After that, in step a82, air-fuel ratio feedback characteristic values are set from these average values.

For example, if there is a significant difference between each mean with DTLR and DTLR, τ _LR In the case of τ _RL , since the correction coefficient median Kc shifts to the lean side or the rich side, the response delay time DLYRL, DLYLR, proportional gain P _RL , P _LR , integral gain I _RL or I _LR depending on the difference of these average values. To modify the value. Store the modified value in memory.

As a result, the shift amount of the median value Kc approaches zero and is corrected to the theoretical air-fuel ratio side. In this way, the response delay time, the proportional gain, and the integral gain are corrected (not necessarily all of the air-fuel ratio feedback characteristic values need to be modified). Next, in step a83, the delay calculation completion plug is set and the drive is carried out in step a68. The distance data OD is input, and at step a56, the next reference value rewriting distance ODX is returned to step a5, for example, as ODX + 800 (mile).

Thereafter, after returning to step a5 of FIG. 10a, if the fuel cutoff area is not, a "no" root is taken in step a6, and after the processing of steps a7 to a9, the mode is temporarily in steps a10 to a12. If YES, it is determined in step a13 'whether or not the plug set of the FB characteristic value calculation is completed. In step a83 of FIG. Perform routine walk during operation. In this case, the air-fuel ratio feedback characteristic values (DLYRL, DLYLR, P _RL , P), which are based on the reference value Vfc for the rich and lean determinations updated as described above and modified based on the response times τ _RL and τ _LR as necessary, are also used. _Air- fuel ratio control is performed based on _{LR and} ).

Thus to each output of the upstream O ₂ sensor 17 outputs Vf and rich-lean judgment upstream O ₂ sensor 17 and the downstream O ₂ sensor 18, the reference value Vfc prepared to be compared in the Vf, changed in reference to Vr In addition, since the air-fuel ratio feedback characteristic value is corrected according to the response time of the upstream O ₂ detector 17, the control accuracy does not change even with the characteristic imbalance and the secular variation of the characteristics of the O ₂ detector. In addition, the exhaust gas purification efficiency by the catalyst switch 9 can also be maintained high, resulting in high control reliability as in the first embodiment described above.

In addition, even when the EGR is not performed or even when the EGR is temporarily performed, even when the rate is low, a good exhaust gas level is obtained, so that the EGR system can be simplified, and the exhaust gas does not sacrifice power performance or operability.

Further, according to the first embodiment, when the response times τ _RL and τ _LR are almost the same (1τ _RL −τ _LR 1 The effect is large when the difference between the response time τ _RL and τ _LR is large (1τ _RL −τ _LR 1> 10 m sec) in this second embodiment for the large effect.

In addition, the rich-lean determination voltage Vfc and the corrected air-fuel ratio feedback characteristic values DLYRL, DLYRL, P _RL , P _LR , I _RL and I _LR are stored in the BURAM 33 and only the ignition switch 26 is turned off. The memory values are not erased. However, when the storage battery 24 is removed, the contents of these memories are also erased. Therefore, if there is a history of detachment of the battery in step (a2) of FIG. 10a, the typical Vf values (DLYRL, DLYLR, Temporarily input the values of P _RL , P _LR , I _RL , and I _LR as initial values, then set the O ₂ detector calibration check completion plug set in step (a71), and set the FB characteristic calculation completion plug set in step (a71-2). Reset is performed.

When the O ₂ detector calibration check completion plug and the FB characteristic value calculation completion plug set are respectively reset, a "no" route is taken in step (a13 ') to satisfy the baseline rewrite condition and perform preprocessing of the baseline rewrite. It is recommended to replace the determination reference value Vfc or to obtain the response time τ _RL , τ _LR and to modify some or all of the air-fuel ratio feedback characteristic values (DLYRL, DLYLR, P _RL , P _LR , I _RL , I _LR ) at these response times. Is done. The processing in this case is the same as in the case of rewriting the electric reference value and modifying the air-fuel ratio feedback characteristics, so that detailed description thereof will be omitted.

3 to 31 show an air-fuel ratio control apparatus for an internal combustion engine as a ninth embodiment of the present invention, and FIG. 13 is a main control block diagram, FIGS. 14A to 14D are flowcharts for explaining the main routine of the mode. Fig. 16 is a flow chart for calculating the deviation between the downstream O ₂ detector output and the target value, and Fig. 16 is a flow chart for correcting the response delay time based on the deviation value obtained in Fig. 15, and Fig. 17 is based on the deviation value obtained in Fig. 15. Flow chart for correcting the air-fuel ratio feedback integral gain, FIG. 18 is a flowchart for correcting the air-fuel ratio feedback proportional gain based on the deviation value obtained in FIG. 15, and FIG. 19 and 20 are graphs for explaining the response delay time correction amount. 27 and 28 are all graphs for explaining the correction method using the air-fuel ratio feedback integral gain, and FIG. 31 is a graph showing the Vf-Vr characteristics. The same reference numerals as those in FIGS. 1 to 12 in the drawings indicate the same parts or processes.

By the way. The air-fuel ratio control device of this third embodiment measures the output Vr of the downstream O ₂ detector 18 during air-fuel ratio feedback in addition to the first embodiment described above, and responds to the delay time DLYRL, DLYLR, and proportional gain P _RL based on this value Vr. , P _LR , integral gain I _RL or I _LR .

However, in the third embodiment, the air-fuel ratio feedback control using the O ₂ detector also includes an output Vf from the upstream O ₂ detector 17 and a useful reference value Vfc (this reference value Vfc is equal to the high level output of the upstream O ₂ detector 17). The intermediate value with the low level output is selected and functions as a so-called rich and lean determination voltage). It is supposed to lean at Vf.

For this reason, the O ₂ detector feedback correction means 37 has an output Vf and rich from the rich lean determination voltage setting means 45 and the upstream O ₂ detector 17 that set the reference value Vfc as shown in FIG. Lean determination voltage setting means 46 and correction coefficient determining means 47 "for determining the air-fuel ratio correction coefficient K _AP in accordance with the comparison result from the comparison means 45. Is different, and the reference value changing means for changing the reference value (rich lean determination voltage) Vfc, for example, for each required travel distance, based on the outputs Vf and Vr from the upstream O ₂ detector 17 and the downstream O ₂ detector 18. Has 48.

The correction coefficient determining means 47 " determines the response self-times DLYRL, DLYRL, proportional gains P _RL , P _LR , integral gain I _RL , based on the output Vr of the downstream O ₂ detector 18 measured during air-fuel ratio feedback. It has a means to calibrate any one of the I _LRs .

Also, the response delay time DLYRL, DLYLR, proportional gain P _RL , P _LR , integral gain I _RL , corrected based on the V _r -V _f characteristic or the reference value Vfc, Vrc or the output Vr of the downstream O ₂ detector 18. I _LR is stored in the BURAM 33.

By the way, although the main routine for changing and updating the said rich lean determination reference value Vfc for every predetermined | prescribed driving distance or after a battery removal history is as shown to FIGS. 14A-14D, these flows are 4A-4D. Since the process is almost the same as shown in Figs. 4A to 4D, the same step numbers are given, and the description thereof is omitted. The baseline rewrite distance DOX is also battery backed. 14A to 14D, the steps different from those of FIGS. 4A to 4D are different from those of steps a70, (a23-2), (a16-4 '), (a19-1') and 14d of FIG. Step a66 '. In addition, step a23-2 is the same as that of the second embodiment.

First, in addition to step (a70) In Vfc (DLYRL, DLYLR), (I _RL, I _LR), to modify by reference to downstream sensor output of the (P _RL, P _LR)), the downstream O ₂ sensor outputs a reference value Vrc Enter initial values for each. Here, this reference value Vrc is determined as follows. That is, as shown in FIG. 31, the output value of the downstream O ₂ detector 18 corresponding to the point almost centered outside the range where dvr / dvf is greater than a certain slope (see step a65 in FIG. 14d) is determined as this reference value Vrc. will be. The value is, for example, Vfc of 0.6 volts.

In addition, in FIG. 31, when Vrc is set at point a, CO purification efficiency is deteriorated. On the contrary, when Vrc is set at β point, NOx purification efficiency is deteriorated. Therefore, Vrc is set at point r in the center as described above.

In addition, the center value in the range of Vf obtained in step a65 in step a66 is set to Vfc, and Vr corresponding to this Vfc is set to Vrc. This Vrc is the output value Vr of the downstream O ₂ detector 18 corresponding to the point γ described above.

Thus, the downstream O ₂ detector output reference value Vrc (this Vrc is obtained as Vfc as described above) in addition to the rich-lean determination reference value Vfc to be compared with the output Vf of the upstream O ₂ detector 17 of the third embodiment. It is updated every mileage or after the battery removal history. In other words, these values Vfc and Vrc are not fixed but set as being variable.

Incidentally, steps a16-4 'and (a19-1') are the same as those shown in the second embodiment, so that the delay time DLYRL has elapsed after Vfc > Vf, respectively, and Vfc. It is to judge whether the delay time DLYRL has elapsed after Vf. However, DLYRL and DLYLR of this third embodiment differ from those of the second embodiment.

Next, a method for correcting response delay times DLYRL, DLYLR, proportional gains P _RL , P _LR , integral gains I _RL and I _LR based on the downstream detector output Vr and reference value Vrc is described.

First, as shown in FIG. 15, in step e1, the output of the upstream O ₂ detector 17 and the outputs IOZSNS (Vf) and IOZCCR (Vr) of the downstream O ₂ detector 18 are read. This read timing is read out every 5 m sec or 10 m sec, for example. In step e2, it is determined whether the upstream O ₂ detector 17 and the downstream O ₂ detector 18 are also active as the output voltage value.

The reference voltage value for this determination is set separately from the upstream O ₂ detector 17 and the downstream O ₂ detector 18.

If both O ₂ detectors 17 and 18 are active, it is determined whether or not the air-fuel ratio feedback is being given to step e3. If yes, it is determined whether the time elapsed after entering the air-fuel ratio feedback mode in step e4. If yes, in step e5, the output frequency IATR of the air flow detector 11, i.e., the amount of intake air, exceeds the set value. Determine whether there are many.

Here, two kinds of set values are prepared, a first set value (XAFSFH) and a second set value (XAFSFL), which are determined using these different set values as the air flow detector output increases and decreases. . That is, since the hysteresis is set in the determination in this step e5, it is advantageous to prevent hunting.

The low intake air operating condition (no-load operation, etc.) In Since the response of the O ₂ sensor late, the output characteristic of the O ₂ sensor are different would be to a judgment such as the step (e5), the air-flow is less than the detector output frequency is set point In this case, the following correction may be made by the reader. In this case, you will learn twice.

In addition, whether the upstream O ₂ sensor output is inverted from the back, "YES" in the step (e5) following step (e6) is determined. As the reference value Vfc for the rich-lean determination on the output of the upstream O ₂ detector 17, Vfc obtained and updated in the main routines (FIGS. 14A to 14D) is used.

In addition, in the above steps (e2 to e6), "no" is returned.

And, by reference to the step (e6) from the "yes" one en step (e7) upstream O ₂ sensor output an output order value IO ₂ CCR and the downstream O ₂ sensor outputs the average value is already stored in the downstream O ₂ sensor at the time of inversion in the case It performs the updating of the downstream O ₂ sensor. That is, the new downstream O ₂ sensor output average value O ₂ RAVE shown on the left side of the lower expression is obtained by O ₂ RAVE = K ₁ (IO ₂ CCR) + (1-K ₁ ) (OZRAVE). In addition, O ₂ RAVE on the right side of the common sense was updated in step e7 of the last timer interrupt routine and stored in RAM. Last downstream O ₂ sensor mean value data. Where K ₁ is the coefficient set in the ROM data.

In addition, the content of the counter COUNT is reduced by 1 at step e8. Here, the initial value of the counter is set in the ROM data and an arbitrary value between (1 to 255) is set, for example. This initial value is set in the counter in step 11 of the routine which is X reaching FIG. 14A at the time of key switch on.

Then, in the next step e9, it is determined whether or not the number of counters is zero, and if "no" is returned, if it is "yes" (that is, the smoothing process of downstream O ₂ sensor output data is sufficiently performed), step e11. the target output voltage of the downstream O ₂ sensor 18 value O ₂ RTRG output average value of the downstream O ₂ sensor 18 during the rich-lean reversal of (this corresponds to Vrc) and upstream O ₂ sensor (17) O ₂ The deviation value? V between these values is obtained from RAVE. In addition, the initial value of the key switch turns on in O ₂ RTRG such as target output value.

When the deviation value ΔV is obtained in this way, the air-fuel ratio feedback characteristic values, that is, the response delay time, the integral gain, and the proportional gain are corrected using this ΔV.

In addition, the change in the output Vr of the downstream O ₂ sensor 18 in the air-fuel ratio feedback is gradual, so not used in a direct-fuel ratio feedback. Since this output Vr is hard to receive the response time difference from rich to lean or lean to rich, it is used to correct the above air-fuel ratio feedback characteristic value.

First, correction of response delay times DLYRL and DLYLR will be described. As shown in FIG. 16, in step e12, DELTA DELAY corresponding to DELTA V obtained in step e11 of FIG.

Further, if DELTA DELAY has a rich to lean and a lean to rich low, the correction characteristics for the former are as shown in Figs. 19 (a) and (b), and the correction characteristics for the latter are 20 (a), It becomes like (b).

That is, ΔDELAY is given as the sum of {ΔDELAY} _P based on the instantaneous value of ΔV and {ΔDELAY} _I based on the integral value of ΔV, where (ΔDELAY) _{R → L} = {(△ DELAY) _{R → L} } _I + {(△ DELAY) _{R → L} } _P , (△ DELAY) _{L → R} = {(△ DELAY) _{L → R} } _I + {(△ DELAY) _{L → R} } _P . Incidentally, the inclinations GP, GI and dead bands ΔdP and ΔdI shown in Figs. 19 and 20 are set in the ROM data.

In this way, after the DELTA DELAY is obtained, new DLTRL and DLTRL are obtained by adding the DELTA DELAY to the reference values (DLYRL) _O and (DLYLR) _O of DLYRL and DLYLR in step e13.

In the next step e14, it is determined whether it is DLYRL or DLYLR. If YES, the result of subtracting DLYLR from DLYRL in step e15 is newly DLYRL. In the next step e16, DLYRL > It is set to ROM data). In addition, while DLYRL does not become this limit value, step e17 is jumped and DLYLR is returned to 0 at step e18. When the DLYRL becomes the limit value, the process of step e18 is performed using the limit value as the DLYRL in step e17.

On the other hand, if DLYRL <DLYRL in step e14, the result of subtracting DLYRL from DLYRL in step e19 is newly DLYLR. In the next step e20, DLYLR> DLYLMT (restriction limit; this is set in ROM data). Determine. While DLYLR does not become this limit value, jump to step e21 and return to step e22 with DLYRL set to zero. When the DLYLR becomes the limit value, the limit value is processed to step e22 as the DLYLR in step e21.

In addition, each delay limit compared in steps e16 and e20 may be the same value or another value.

DLYRL and DLYLR are each backed up by a battery, but the initial value in step a70 "

Thus, when DLYRL and DLYLR are corrected and enriched based on the downstream O ₂ detector output, DLLR is added as shown in Figs. 25A to 25E, and DLYRL as shown in Figs. 26A to 26C when leaning. Adding is done.

In this manner, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured every time (or each time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc), and the moving average value is determined by Vrc. Since the response delay time is corrected so as to be equivalent, the air-fuel ratio control can be performed with higher reliability and precision except that effects or advantages almost similar to those of the above-described embodiments can be obtained.

Next, the correction of the air-fuel ratio feedback integral gains I _RL and I _LR will be described. As shown in FIG. 17, in step e23, DELTA I corresponding to DELTA V obtained in step e11 of FIG.

In addition, there are those from lean to lean and those from lean to lean. The correction characteristics for the former are as shown in Figs. It becomes like (b). That is, ΔI is considered as the sum of {ΔI} _P based on the instantaneous value of ΔV and {ΔI} _I based on the integral value of ΔV,

(ΔI) _{R → L} = {(ΔI) _{R → L} } _I + {(ΔI) _{R → L} } _P

(ΔI) _{L → R} = {(ΔI) _{L → R} } _I + {(ΔI) _{L → R} } _P

The functional relationships (tilt or dead band) shown in these 21 and 22 degrees are set in the ROM data.

Thus, after obtaining? I, new I _RL and I _LR are obtained by adding these? I to the reference values I _RLO and I _LRO of I _RL and I _LR in step e24.

In the next step (e25), it is determined whether I _RL &gt; I _F (upper limit: this value is set in the ROM data). If " no &quot;, then in step e27, I _RL &lt; I _L (lower limit: This value is set in the ROM data).

If YES in step e25, I _F is I _RL in step e26. If YES is again in step e27, I _L is I _RL in step e28.

Also, in the case of "no" in step e27 or after the processing of steps e26 and e28, in the next step e29, is I _RL > I _H (upper limit: this value is set in the ROM data)? If NO, in step e31, it is determined whether I _LR &lt; I _L (lower limit: this value is set in the ROM data).

If YES in step e29, I _{H is set} to I _LR in step e30. If YES is returned in step e31, I _L is returned as I _LR in step e32.

In addition, each upper limit compared with step e25 and e29 may be the same value, or a different value may be sufficient, and the lower limit compared with step e27 and e31 may be the same value or a different value.

In addition, the integral gains I _RL and I _LR are battery-backed up respectively.

Thus, when I _RL and I _LR are corrected and enriched based on the downstream O ₂ detector output Vr, as shown in Figs. 27A to 27C, I _RL is made larger and I _LR is made smaller.

In this way, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured at a predetermined time (or each time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc) and its moving average is equal to Vrc. Since the integral gain is corrected so as to achieve almost the same effects and advantages as the above embodiments, the air-fuel ratio control can be performed with higher reliability and precision.

Next, the correction of the air-fuel ratio feedback proportional gains P _RL and P _LR will be described. As shown in FIG. 18, in step e33, DELTA P according to DELTA V obtained in step e11 of FIG.

Incidentally, this ΔP has a rich to lean and a lean to rich furnace, and the correction characteristics for the former are as shown in FIGS. 23a and 23b, and the correction characteristics for the latter are as shown in 24a and 24b. That is, ΔP is given as the sum of {ΔP} _P based on the instantaneous value of ΔV and {ΔP} _I based on the integral value of ΔV,

(ΔP) _{R → L} = {(ΔP) _{R → L} } _P + {(ΔP) _{R → L} } _P

(ΔP) _{L → R} = {(ΔP) _{L → R} } _P + {(ΔP) _{L → R} } _P

The function relationships (tilt and dead band) shown in these 23 and 24 degrees are set in the ROM data.

Thus obtained by the △ P is obtained following a step (e34) and the new _RL P, P _LR is hameuroseo these _RL △ P to P, the reference value of the P _{LR RLO} P, P _LRO.

In the next step e35, P _RL &gt; P _R (upper limit value: this value is set in the ROM data).

Thus obtained to the next △ P calculate the new _RL P, P _LR is the hameuroseo thereof △ P to P _RL, the reference value of the P _{LR RLO} P, P _LRO in step (e34).

In the next step e35, it is determined whether P _RL &_gt; P _H (upper limit: this value is set in the ROM data), and if " No &quot;, in step e37, P _RL &lt; P _L (lower limit value). This value is set in the ROM data).

If "Y" in step e35, P _H is P _RL in step e36, and if "Yes" in step e37, P _L is P _RL in step e38.

In the case of "no" in step e37 or after the processing of steps e36 and e38, P _LR > P _H (upper limit value: this value is set in the ROM data) in the next step e39. If NO, in step e41, it is determined whether P _LR &lt; P _L (lower limit: this value is set in ROM data).

If the if the "Yes" at the step (e39) and the P _H in step (e40) to the P _LR also "yes" in step (e41) to return to the P _L at the step (e42) to the P _LR.

In addition, each upper limit value compared in step e35 and e39 may be the same value, and the lower limit value compared in step e37 and e41 may be the same value, or a different value. In this way, P _RL and P _LR are corrected based on the downstream O ₂ detector output Vr, and in the case of enriching, P _RL is made larger and P _LR is made smaller as shown in FIGS. 29a and 29c. .

In this manner, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured at a predetermined time (or each time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc), and the moving average value is determined by Vrc. Since the proportional gain is corrected to be equivalent, the effects or advantages almost similar to those of the above embodiments can be obtained, and the air-fuel ratio control can be performed with high reliability and precision.

In the third embodiment, only a part of the response delay time, the integral gain, and the proportional gain may be corrected so that the moving average value of the output Vr of the downstream O ₂ detector 18 becomes equal to Vrc.

32 to 53 show an air-fuel ratio control apparatus for an internal combustion engine as a fourth embodiment of the present invention, FIG. 32 is a main control block diagram, and FIGS. 33A to 33D are flowcharts for explaining the main routines thereof; FIG. 34 is a flowchart for calculating a deviation value between the downstream O ₂ detector output and a target value (reference value), FIG. 35 is a flowchart for correcting a response lag time based on the deviation value obtained in FIG. 34, and FIG. A flowchart for correcting the air-fuel ratio feedback integral gain based on the deviation value obtained at 34 degrees, FIG. 37 is a flowchart for correcting the air-fuel ratio feedback proportional gain based on the deviation value obtained at 34 degrees, and FIG. 38 is a deviation value obtained in FIG. by reference to both the flow chart, the turns 39 and 40 for correcting the reference value for the rich-lean judgment to be compared with the upstream O ₂ sensor output illustrating the response delay time correction A graph, 41, and 42 degrees is the air-fuel ratio feedback integrating both the gain correction amount described below graph, the 43 and 44 degrees for the comparison and output the upstream O ₂ sensor graphs, turning claim 45 for explaining the air-fuel ratio feedback proportional gain correction amount Graphs for explaining the correction amount of the reference value for the rich and lean determination, 46 and 47 are all graphs for explaining the response delay time correction method, and FIGS. 48 and 49 are all corrections by the air-fuel ratio feedback integration gain Graph for explaining the method, 50 and 51 are all graphs for explaining the correction method by the air-fuel ratio feedback proportional gain, Figure 50 and 51 are all graphs for explaining the correction method by the air-fuel ratio feedback proportional gain , Figs. 52 and 53 are graphs for explaining the correction method by the reference value for the rich-lean determination to be compared with the upstream O ₂ detector output. In the drawings, the same reference numerals as those in FIGS.

By the way. The air-fuel ratio control device of this fourth embodiment measures the output Vr of the downstream O ₂ detector 18 during air-fuel ratio feedback and based on this value Vr, response delay times DLYRL, DLYLR, proportional gains P _RL , P _LR , and integral gain I _RL , is referred to in addition to, the rich to be compared with the upstream O ₂ sensor output, if necessary. Lin determination reference value Vfc (below, this embodiment for correcting either of the I _LR, O2RLL the rich Lin determination reference value Vfc for ) Is also corrected.

However, the high level output of the fourth exemplary examples, when the air-fuel ratio feedback control writing the O ₂ sensor has a reference value of the output Vf, and the time taken from the upstream O ₂ sensor 17 O2RLL (the reference value O2RLL the upstream O ₂ sensor 17 The intermediate value between the low level output and the low level output is selected and functions as a so-called rich / lean determination voltage). It is supposed to lean at Vf.

For this reason, the O ₂ detector feedback correction means 37 includes the outputs Vf from the rich-lean determination voltage means 45 'and the upstream O ₂ detector 17 that set the reference value O ₂ RLL as shown in FIG. Comparing means 46 for comparing with reference value O2RLL from rich-lean determination voltage setting means 45 ', and correction coefficient determining means for determining the air-fuel ratio correction coefficient K _AF in accordance with the comparison result from this comparing means 46. with (47 "), and also takes the case of the air-fuel ratio control apparatus is output from the upstream sensor 17 and the downstream O ₂ sensor (18) Vf, the reference value Vrc for by reference to Vr downstream O ₂ sensor, for example, It has the reference value changing means 50 which can be changed for every traveling distance.

And, there is a reference value changing unit 50 is equipped with a characteristic calculation means 49 for calculating a characteristic between the two output of the upstream O ₂ sensor 17 and the downstream O ₂ sensor 18 like that illustrated in 8b FIG. It is supposed to be updated as the reference value Vrc for the downstream O ₂ detector 18 obtained by this characteristic calculating means 49. It is assumed that the reference value setting means 51 has this update function.

The reference value for the downstream O ₂ sensor from a reference value setting means (51) Vrc signal is rich, there is input to the lean judgment voltage means (45 ') and the correction factor determining means (47) rich. Lin determined voltage determining means (45 ') and the correction factor determining means (47') is by reference to the comparison result to the output Vr from the reference value Vrc and the downstream O ₂ sensor 18 for each of the downstream O ₂ sensor air-fuel ratio control by the air-fuel ratio control means Also has a function of air-fuel ratio control correction means 45'A (47 "A) which applies correction to the air-fuel ratio, that is, air-fuel ratio control correction means 45'A in the rich-lean determination voltage setting means 45 '. ) Can correct the reference value O2RLL for the rich-lean determination based on the deviation value? V between the reference value Vrc for the downstream O ₂ detector and the output Vr of the downstream O ₂ detector 18 measured during the air-fuel ratio feedback. Air-fuel ratio control correction means in the means 47 (47 "A), the downstream O _2, based on the deviation values △ V between the output Vr of the detector reference value Vrc as the downstream O ₂ sensor 18 measuring the air-fuel ratio feedback for the response delay time DLYRL, DLYLR, proportional gain P _RL, P _LR , integral gain I _RL or I _LR can be corrected.

The Vr-Vf characteristic, the reference value Vrc, or the rich-lean determination voltage O2RLL, the response delay time DLYRL, DLYLR, the proportional gains P _RL , P _LR , the integral gains I _RL , I _LR are stored in the BURAM 33.

By the way, the main routine for changing and updating the reference value Vrc every predetermined driving distance or after the battery detachment history is shown in FIGS. 33A to 33D, and the flow thereof is almost the same as shown in FIGS. 14A to 14D. Therefore, step numbers like 14a to 14d are assigned, and description thereof is omitted. In addition, the reference value rewrite distance DOX battery is backed up. Incidentally, in Figs. 33A to 33D, steps different from Figs. 14A to 14D are steps a14 ', (a16-4'), (a19-1 ") and (a70 ') of Fig. 33A.

First, the step (a14 ') In this and the reference value to be compared with the output Vf of the upstream O ₂ sensor 17 is in O2RLL as described above, the step in step (a16-4 ") and step (a19-1") ( a2 'is O2RLL. By the time the air-fuel ratio feedback control using the detector as it compares with a reference value of the output Vf, and the time taken from the upstream O ₂ sensor (17) O2RLL (rich-lean judgment voltage) when O2RLL> Vf yen rich Chemistry and Anti O2RLL It is supposed to lean at Vf.

In step (a70 '), Vfc, (DLYRL, DLYRL), (I _RL , I _LR ), and (P _RL , P _LR ) are adjusted as necessary except for the modification based on the downstream O ₂ detector output. Initial values are respectively input for the judgment reference value O2RLL. Here, the reference value O2RLL is set at, for example, about 0.6 volts.

Next, correct the response delay times DLYRL, DLYLR, proportional gains P _RL , P _LR , integral gain I _RL , I _LR , and reference value O ₂ RLL for the rich-lean determination based on this downstream O ₂ detector output Vr and reference value Vrc. Explain how to do this.

First, the target output voltage value of the downstream O ₂ sensor 18 O2TRG output average O2RAVE the downstream sensor 18 during the rich-lean reversal of (this corresponds to Vrc) and upstream O ₂ sensor 17, that is, the downstream O _The flow chart for calculating the deviation value? V of these values from the reference value Vrc for the _two detectors and the output Vr of the downstream O ₂ detector 18 is as shown in FIG. 15, and is shown in FIG.

When the deviation value? V is obtained in this way, the response delay time, the integral gain, the proportional gain, and the rich and lean judgment reference value are corrected using this? V.

In addition, the change in the output Vr of the downstream O ₂ sensor 18 in the air-fuel ratio feedback is gradual, so not used in a direct-fuel ratio feedback. Since this output Vr is hard to receive the response time difference from the rich to the lean or the lean to the rich, it is used to correct the response delay time, the integral gain, the proportional gain, and the reference value for the rich and lean determination as described above.

First, the flowcharts for correcting the response delay times DLYRL and DLYRL are the same as those shown in FIG.

In the step e15 of FIG. 35, ΔDELAY obtained according to ΔV includes those of the rich to the lean and the lean to the rich. The correction characteristics for the former are the same as those of FIG. 19 and 39. The correction characteristic for is shown in FIG. 20 and 40. Also, the functional relationships (tilt and dead band) shown in these 39th and 40th degrees are also set in the ROM data.

In this case, the DLYRL and DLYLR are calibrated based on the downstream O ₂ detector output and are enriched in the same manner as shown in FIGS. 25A to 25C, and the DLYLR is added and rinsed as shown in FIGS. 46A to 46C. The case is the same as the case shown in FIGS. 26A to 26C, and the addition of DLYRL is performed as shown in FIGS. 47A to 47C.

In this way, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured every time (every time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc) and the moving average value is equal to Vrc. Since the response delay time is corrected so as to correct the air-fuel ratio control, the air-fuel ratio control can be performed with high reliability and precision.

Next, a flowchart for correcting the air-fuel ratio feedback integration gains I _RL and I _LR is the same as that shown in FIG.

In the step e23 of FIG. 36, ΔI obtained according to ΔV includes those from rich to lean and from lean to rich, and the correction characteristics for the former are the same as those of FIG. 21 and 41. The correction characteristic for the latter is shown in FIG. 22 and is shown in FIG. 42. In addition, the functional relationships (tilts and dead bands) shown in Figs. 41 and 42 are set in the ROM data.

In this case, the correction and enrichment of I _RL and I _LR based on the O ₂ detector output is made as in the case of Figs. 27a to 27c, and the I _RL is made smaller as shown in Figs. 48a to 48c. In the case of increasing the I _LR and leaning, as shown in FIGS. 28A to 28C, as shown in FIGS. 49A to 49C, the I _RL is increased and the I _LR is decreased.

In this way, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured every predetermined time (or each time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc) and its moving average is equal to Vrc. Since the integral gain is corrected so that the air-fuel ratio control is corrected, the same effects or advantages are obtained as in the case of correcting the response delay time described above.

Next, the flow chart for the correction of the air-fuel ratio feedback proportional gains P _RL and P _LR is as shown in FIG. 18 and is shown in FIG. 37.

Incidentally, in the step (e33) of FIG. 37, DELTA P obtained in accordance with DELTA V includes rich to lean and lean to rich. The correction characteristics for the former are the same as in FIG. The correction characteristic for the latter is shown in FIG. 24 and shown in FIG. 44. In addition, the functional relationships (tilt and dead band) shown in FIGS. 43 and 44 are set in the ROM data.

And as such if the screen corrected by reference to P _RL, P _LR downstream sensor output and rich has the the same as the case shown in 29a to 29c even with the box Claim 51a to 51c As shown in Figure the P _RL zoom P _{Making LR} small is done.

In this manner, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured at a predetermined time (or each time the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc), and the moving average value is determined by Vrc. Since the proportional gain is corrected to be equal and correction is added to the air-fuel ratio control, the same effect or advantage is obtained in this case as in the case of correcting the response delay time or the integral gain described above.

Next, correction | amendment of the reference value O2RLL for rich. Lean determination is demonstrated. First, as shown in FIG. 38, in step e43,? O2RLL different from? V obtained in step e11 in FIG.34 is calculated.

The correction characteristics for the ΔO ₂ RLL are as shown in the 45th (a) and (b) degrees. That is, ΔO2RLL is given as the sum of (ΔO2RLL) P based on the net value of ΔV and (ΔO2RLL) I based on the integral value of ΔV,

DELTA O2RLL = (ΔO2RLL) I + (ΔO2RLL) P.

The functional relationship (tilt G or deadband? D) shown in FIG. 45 is also set in the ROM data. In this way, after obtaining? O2RLL, a new O2RLL is obtained by applying these? O2RLL to the reference value (O2RLL) O of O2RLL in step (e44).

In the next step e45, it is determined whether O2RLL> XO2H (upper limit: this value is set in ROM data). If "No", in step e47, O2RLL <XO2L (lower limit: this value is ROM data). Is set in step 1).

If, If "Yes", in addition, step (e47) as a O2RLL XO2H in step (e46) if the "Yes" at the step (e45), and the XO2L in step (e48) to O ₂ RLL.

In the case of NO in step e47 or after the processing of steps e46 and e48, the process returns.

As described above, when the O ₂ RLL is corrected based on the downstream O ₂ detector output Vr and enriched, as shown in FIGS. 52 (a) to 52 (c), when the O ₂ RLL is enlarged and leaned, the 53rd to 53c As shown in the figure, reducing the O ₂ RLL is performed.

In this way, the output Vr of the downstream O ₂ detector 18 during the air-fuel ratio feedback is measured at regular intervals (or as the output Vf of the upstream O ₂ detector 17 crosses the reference value Vfc) and its moving average is equal to Vrc. Since the reference value for the rich and lean determination is corrected as much as possible to correct the air-fuel ratio control, the same effects or advantages as in the case of correcting the response delay time, integral gain or proportional gain described above are obtained.

In addition, in the fourth embodiment, only a part of the response delay time, the integral gain, the proportional gain, and the Rich and Lean determination reference values so that the moving average value of the output Vr of the downstream O ₂ detector 18 becomes equal to Vrc. You may correct.

Further, in the third and fourth embodiments, the output of the downstream O ₂ detector is output whenever the output of the upstream O ₂ detector is reversed in steps e6 and e7 of the flowcharts of FIGS. 15 and 34. Although the average value O ₂ RAVE has been updated, this update may be performed whenever the intake air amount reaches a predetermined value (that is, every time the integrated value of the intake air amount data reaches the predetermined value). As shown in FIG. 2, when discrete pulses of the frequency corresponding to the intake air are input from the air flow detector (Kalman vortex flowmeter) 11 to the ECU 23, FIG. 55 is used instead of the flowchart of FIG. 15 or 34. And the flow chart of FIG. 56. That is, as shown in FIG. 55, in the routine that operates each time a pulse synchronized with the occurrence of Kalman arrives, a step of integrating the number of pulses is provided and a step of the timer embedding routine shown in FIG. In the case of (e60), it is determined whether or not the accumulated value exceeds OX, and when it exceeds (after resetting the integrated value data Qa), the same as in the step e7 described above in the step e7 '. The downstream O ₂ detector output average value is updated. (In addition, if there is a determination of "no" in any one of steps (e2) to (e5) in FIG. 56, the integrated value data Qa is reset to O.) In Fig. 56, the same reference numerals as in Figs. 15 and 34 in the steps of Fig. 56 and the same reference numerals as in Figs. 15 and 34 indicate the same contents as those in Figs. 15 and 34).

In addition, in this FIG. 56, step e5 may be abbreviate | omitted.

FIG. 54 is an overall configuration diagram showing an engine system having an air-fuel ratio control apparatus for an internal combustion engine as a fifth embodiment of the present invention, in which the same reference numerals as those in FIG. 3 show the same parts.

In the fifth embodiment, the secondary air supply passage 60 is connected to the upstream side of the catalytic converter 9, and the electronic control valve 61 is retrofitted to the secondary air supply passage 60, and the downstream O ₂ outputs Vr and the reference value Vrc wareul compared to the comparison result to therefore control valve 61 for opening or chungryeok coefficient by changing the second would be to control the air-fuel ratio by adjusting the supply amount of the air amount the downstream O ₂ sensor (18 of the sensor 18 The reference value Vrc is adjusted based on the output of the upstream O ₂ detector 17 and the output of the downstream O ₂ detector 18.

In other words, each output state of the upstream O ₂ detector 17 and the downstream O ₂ detector 18 is sampled while artificially changing the secondary air supply amount, as shown in Fig. 8C (Fig. 8B). The correlation graph is obtained and the result is the reference value Vrc for the downstream O ₂ detector.

In addition, the correction method of the reference value in such a case is almost the same as shown in FIG. 38, 52, 53 degrees.

In Fig. 54, P denotes an electric pump and F denotes an air filter.

In the second embodiment described above, the processing in step e23 of the main routine may be K _C = k K _C + (1-K) K _FB .

0 here K Although 1, 0 <K <1 may be sufficient.

In the first embodiment described above, a flow for obtaining K _C may be added as in the second embodiment.

In addition, in the above-described first embodiment, response delay times DLYLR and DLYRL may be considered. In that case, for example, the processing corresponding to step a16-4 'and step a1a-1' in Fig. 11A is placed in Fig. 4A.

Also, it performed at the downstream O ₂ sensor 18, the role of the upstream O ₂ sensor 17 in each of the above embodiments, and may perform the role of the downstream O ₂ sensor 18 at the upstream O ₂ sensor 17 do.

In addition, fail-safe (the back of the fail-safe (fail safe) (backup) as writes, the downstream O ₂ sensor upstream sensor (17) to (18) of the downstream O ₂ sensor 18, the upstream O ₂ sensor 17 You may write for up).

In addition, the present invention can be adopted throughout the system for performing feedback control of the O ₂ detector, or, of course, also applies to an SPI (single point fuel injection) engine system outside the MPI engine system.

As described above, according to the air-fuel ratio control apparatus of the internal combustion engine described in claim (1) related to the present invention, the oxygen concentration in the exhaust gas is provided upstream of the catalytic converter for exhaust gas purification installed in the exhaust machine of the internal combustion engine. detection of the from the second the first O ₂ sensor in addition to matching the O ₂ sensor of the above to detect the O ₂ sensor and the oxygen concentration in the catalytic converter, the exhaust gas is installed in or downstream of the first that detects the An air-fuel ratio control means for controlling the air-fuel ratio of the internal combustion engine on the basis of a comparison result between the value and the required reference value, and the reference value is determined based on the outputs from the first O ₂ detector and the second O ₂ detector. since the reference value changing means is provided for changing the rich Lin determination reference value to be compared with the output of the upstream O ₂ sensor of the catalytic converter upstream O ₂ detected And it can be changed by reference to each of the output from the downstream O ₂ sensor, which the control accuracy by the aging change of the product attributes fire gyunjeong or characteristics of each of the O ₂ sensor is not changed Besides exhaust gas purification efficiency by the catalytic converter, FIG. There is an advantage that it can be kept high and as a result high reliability is obtained.

In addition, according to the air-fuel ratio control apparatus for an internal combustion engine according to the present invention be installed in the purification of exhaust gas it is installed in the ship machinery of the internal combustion engine catalytic converter upstream of the the O ₂ sensor of claim 1 for detecting the oxygen concentration in the exhaust gas the catalyst is installed in or downstream of the converter during the addition to matching the O ₂ sensor of claim 2 for detecting the oxygen concentration in the exhaust gas first of the O ₂ sensor and an O ₂ sensor in the second one of O ₂ An air-fuel ratio control means for controlling the air-fuel ratio of the internal combustion engine on the basis of a comparison result between the detection value from the detector and the required reference value, and based on the outputs from the first O ₂ detector and the second O ₂ detector. it is set by the second reference value setting means and the same second reference value setting means for setting a second reference value for the O ₂ sensor on the other side of the O ₂ sensor and the second O ₂ sensor in the first of said Second reference value and the first of the O ₂ sensor and a second of the O ₂ sensor by reference to the comparison result of the output from the other side of the O ₂ sensor of applying a correction for the air-fuel ratio control by the air-fuel ratio control means for the air-fuel ratio control The above-described effects and advantages are obtained by providing the air-fuel ratio control correction means for applying correction to the air-fuel ratio control by the means.

Claims (1)

A first oxygen concentration detector installed upstream of the exhaust gas purifying catalyst converter installed in the exhaust system of the internal combustion engine and detecting oxygen concentration in the exhaust gas, and an oxygen concentration in the exhaust gas installed inside or downstream of the catalytic converter. And a second oxygen concentration detector for detecting a value, and comparing the detected value from one of the first oxygen concentration detectors and the second oxygen concentration sensor with a predetermined reference value. An air-fuel ratio control means for controlling the air-fuel ratio of the internal combustion engine on the basis of the first oxygen concentration sensor and the second oxygen concentration sensor based on the outputs from the first oxygen concentration sensor and the second oxygen concentration sensor; Second reference value setting means for setting a second reference value for the other oxygen concentration sensor in the concentration sensor, and a second reference value set by the second reference value setting means The air-fuel ratio control correction means for correcting the air-fuel ratio control by the air-fuel ratio control means on the basis of a comparison result between the value and the output from the other oxygen concentration sensor of the first oxygen concentration sensor and the second oxygen concentration sensor. Air-fuel ratio control device of an internal combustion engine, characterized in that installed.