JPH0555704B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an air-fuel ratio control method, and particularly to an air-fuel ratio control method suitable for use in a vehicle internal combustion engine having an electronically controlled fuel injection device. .

[Conventional technology]

The electronically controlled fuel injection system calculates the basic fuel injection time TP based on the engine rotation speed NE detected by the rotation speed sensor and the intake air amount Q detected by the intake air amount sensor, and The final fuel injection time τ is calculated by applying various corrections to the basic fuel injection time TP, and the injection valve is opened for the final fuel injection time τ to inject fuel.

On the other hand, in this type of fuel injection control device that uses a three-way catalytic converter to simultaneously remove CO, HC, and NOx from the exhaust gas as a countermeasure against exhaust emissions, from the viewpoint of efficiently removing the three components mentioned above, It is desired to control the fuel ratio near the stoichiometric air-fuel ratio. Therefore, an oxygen sensor is installed in the exhaust passage, and under certain conditions, the feedback correction coefficient FAF is adjusted so that the air-fuel ratio becomes close to the stoichiometric air-fuel ratio based on the air-fuel ratio signal from the oxygen sensor.
is calculated to perform air-fuel ratio feedback control.

In an electronically controlled fuel injection system that performs air-fuel ratio feedback control, it compensates for differences in air-fuel ratio due to variations between parts, compensates for air-fuel ratio due to high-altitude driving, and compensates for air-fuel ratio differences due to changes in the intake air amount sensor over time. In order to compensate for changes in the air-fuel ratio, the learning correction coefficient FG is calculated by learning the air-fuel ratio under predetermined conditions during the feedback control.

Then, the final fuel injection time τ is, for example, τ=
It is determined by the formula TP x FAF x FG x K. Here, K is a correction coefficient based on water temperature, intake air temperature, etc.

[Problem that the invention seeks to solve]

When learning the air-fuel ratio, the fuel evaporated in the fuel tank and stored in the canister (hereinafter referred to as evaporated fuel) enters the combustion chamber under predetermined conditions, including at least that the throttle valve is not fully closed. It must be taken into account that the air-fuel ratio is temporarily enriched. The effect of such evaporated fuel on the air-fuel ratio is shown in Figure 2, and in extreme cases, the intake air amount Q may be 100 m ³ /
Even in a region with a high air flow rate of about 100 mph, it can become about 10% richer.

Therefore, if you stop driving the vehicle immediately after learning such changes in the air-fuel ratio due to evaporated fuel, the next time you start the vehicle, the air-fuel ratio will be too lean, causing problems such as poor starting performance. .
Therefore, it is necessary not to learn about air-fuel ratios that are rich due to evaporated fuel.

The above-mentioned compensation for the air-fuel ratio at high altitudes means that the air density becomes smaller at higher altitudes, so the air-fuel ratio is prevented from becoming richer at higher altitudes, but the effect of high altitudes on the air-fuel ratio is As shown in Figure 3, it is almost constant regardless of the intake air amount.
Therefore, in areas other than the region where the throttle valve is fully closed, it is difficult to determine whether the reason why the air-fuel ratio becomes rich is due to evaporated fuel or high-altitude driving.

On the other hand, if the intake air amount sensor becomes clogged due to changes over time, as shown in FIG. 4, the area where the intake air amount is smaller has an effect on the air-fuel ratio. Therefore, if the air-fuel ratio differs by, for example, 1.5% or more between the throttle valve fully closed region and the other regions, it is determined that the intake air amount sensor is clogged, and the air-fuel ratio is set to τ (excess air ratio) = 1. In conventional control that subtracts the learning correction coefficient so that
Similar learning is performed in the case of the influence of the air-fuel ratio as shown in the figure, and the compensation of the air-fuel ratio due to changes over time and the compensation of the air-fuel ratio due to evaporated fuel are superimposed, making it difficult to properly compensate for the air-fuel ratio. Furthermore, if the rider descends from a high altitude with the throttle valve fully closed, there is a risk that compensation for clogging may not be performed accurately due to the effects of the high altitude.

[Means action to solve the problem]

The present invention uses the average value as a learning condition for the learning correction coefficient DFC for clogging compensation of an air flow meter, which is learned to increase when the average value FAFAV1 of the feedback correction coefficient FAF is large and to decrease when it is small. It is determined that the judgment value FAFAV2 related to FAFAV1 is within a predetermined range, and the judgment value FAFAV2 approaches the average value FAFAV1 under conditions in which evaporated fuel in the canister is sucked into the intake pipe.

[First Embodiment] FIG. 5 shows an example of an electronically controlled fuel injection type internal combustion engine to which the present invention is applied.
Reference numeral 2 indicates an intake passage, 14 a combustion chamber, and 16 an exhaust passage. An intake air amount sensor (air flow meter) 20 provided in the intake passage 12 upstream of the throttle valve 18 is connected to a control circuit 22 via a signal line l1, and generates a voltage according to the amount of intake air. The intake temperature sensor 21 is provided in the intake passage 12 upstream of the throttle valve 18, and is connected to the signal line l.
2 to the control circuit 22, and generates a voltage according to the intake air temperature. Intake air is taken in through an air cleaner and an intake air amount sensor 20 (not shown), and whose flow rate is controlled by a throttle valve 18 that is linked to an accelerator pedal (not shown).
The air is guided to the combustion chamber 14 of each cylinder via the surge tank 24 and intake valve 25.

A fuel injection valve 26 is provided for each cylinder,
The opening/closing is controlled in response to electrical drive pulses supplied from the control circuit 22 via the signal line l3, and pressurized fuel sent from a fuel supply system (not shown) is fed into the intake passage 12 near the intake valve 25, that is, the intake port. Inject intermittently into the area. The exhaust gas after being combusted in the combustion chamber 14 is discharged into the atmosphere via the exhaust valve 28, the exhaust passage 16, and the three-way catalytic converter 30.

Crank angle sensors 34 and 36 are attached to the distributor 32 of the engine, and these sensors 34 and 36 are connected to the control circuit 22 via signal lines 14 and 15. These sensors 34 and 36 output pulse signals each time the crankshaft rotates 30 degrees and 360 degrees, respectively, and these pulse signals are supplied to the control circuit 22 via signal lines l4 and l5, respectively.

The distributor 32 is connected to an igniter 38, and the igniter 38 is connected to the control circuit 22 via a signal line l6.

Reference numeral 40 denotes an handle switch (LL switch) which is interlocked with the throttle valve 18 and is closed when the throttle valve 18 is fully opened.
It is connected to the control circuit 22 via.

A signal responsive to the oxygen concentration in the exhaust gas is output to the exhaust passage 16. In other words, the air-fuel ratio generates an output voltage in steps with the stoichiometric air-fuel ratio as the boundary.
An O ₂ sensor 42 is provided, and its output signal is connected to the control circuit 22 via a signal line l8.
The three-way catalytic converter 30 uses this O ₂ sensor 42
It is installed downstream of the exhaust gas and simultaneously purifies the three harmful components of exhaust gas: HC, CO, and NOx.

Further, reference numeral 44 detects the engine cooling water temperature,
This is a water temperature sensor that generates a voltage according to the temperature, and is attached to the cylinder block 46 and connected to the control circuit 22 via a signal line 19.

As shown in FIG. 6, the control circuit 22 includes a central processing unit (CPU) 22a that controls various devices;
It has a read-only memory (ROM) 22b in which various numerical values and programs are written in advance, a random access memory (RAM) 22c in which numerical values and flags in the calculation process are written in a predetermined area, and an analog multiplexer function, and can accept analog input signals. A/D converter (ADC) that converts to digital signals
22d, input/output interface (I/O) 22e to which various digital signals are input, input/output interface (I/O) 22f to which various digital signals are output, is supplied with power from the auxiliary power source and retains memory when the engine is stopped. It is composed of a backup memory (BU-RAM) 22g, and a bus line 22h to which each of these devices is connected.

The ROM 22b contains a main processing routine program, a program for calculating fuel injection time (pulse width), a program for calculating an air-fuel ratio feedback correction coefficient, a learning correction coefficient to be described later, and various other programs, as well as programs for calculating these calculations. Various necessary data are stored in advance.

And air flow meter 20, intake temperature sensor 2
1. O ₂ sensor 42 and water temperature sensor 44 are A/D
It is connected to a converter 22d, and voltage signals S1, S2, S3, S4 from each sensor are sequentially converted into binary signals in accordance with instructions from the CPU 22a.

A pulse signal S5 for every 30 degrees of crank angle from the crank angle sensor 34, a pulse signal S6 for every 360 degrees of crank angle from the crank angle sensor 36, and an idle signal S7 from the idle switch 40 are transmitted via the I/O 22e. The signal is taken into the control circuit 22. A binary signal representing the engine speed is formed based on the pulse signal S5, and the pulse signals S5 and S6 work together to generate a request signal for calculating the fuel injection pulse width, an interrupt signal for starting fuel injection, and a cylinder discrimination signal. etc. are formed. Furthermore, it is determined based on the idle signal S7 whether the throttle valve 18 is substantially fully closed.

The I/O 22f outputs a fuel injection signal S8 and an ignition signal S9 formed by various calculations to the fuel injection valves 26a to 26d and the igniter 38, respectively.

The fuel injection time (injection amount) in the internal combustion engine configured as described above is determined, for example, as follows.

τ=TP×FAF×FG×K...(1) Here, τ=Final fuel injection time TP=Basic fuel injection time FAF=Feedback correction coefficient FG=Learning correction coefficient K=Correction coefficient based on water temperature, intake temperature, etc. Basic The fuel injection time TP is determined by reading from a predetermined table or by calculation based on the intake air amount Q and the engine speed NE.

The feedback correction coefficient FAF is a value that increases the injection amount if it is determined that the air-fuel ratio is lean based on the air-fuel ratio signal S3 from the O ₂ sensor 42 under feedback control conditions, for example.
1.05, and if the air-fuel ratio is determined to be rich based on the air-fuel ratio signal S3, it will be a value that reduces the injection amount, for example 0.95, and if it is not under the feedback control condition, the correction coefficient FAF will be 1.0.

An example of the calculation procedure for the feedback correction coefficient FAF is shown in FIG.

In step S1, it is determined whether a feedback condition is satisfied. For example, if the engine water temperature THW is not in the starting state or increasing after starting,
The conditions for feedback control are met when the temperature is 50°C or higher and the power is not being increased. If the conditions for feedback control are not satisfied, step S2
Then, set the feedback correction coefficient FAF to 1.0 so that no feedback control is executed, and end this procedure. If the conditions are met, proceed to step S3
Proceed to. In step S3, the air-fuel ratio signal S3 is read. In step S4-1, an air-fuel ratio lean/rich flag is formed based on the voltage value represented by the air-fuel ratio signal S3 so that it becomes "1" when rich and "0" when lean. If the flag is "1", it is determined that the air-fuel ratio is too rich, and a procedure is executed to make the air-fuel ratio lean.

That is, in step S5, the flag CAFL is set to zero, and the process proceeds to step S6, where it is determined whether the flag CAFR is zero. When it first shifts to the over-concentrated side, a flag is displayed.
Since CAFR is zero, proceed to step S8, and RAM2
A predetermined value α1 is subtracted from the correction coefficient FAF stored in 2b, and the result is set as a new correction coefficient FAF. In step S9, the flag CAFR is set to 1. Therefore, if excessive concentration is determined twice or more in succession in step S4-2, a negative determination is always made in step S6 passed from the second time onwards, and a predetermined value β1 is subtracted from the correction coefficient FAF in step S7. , the FAF calculation is completed using the result as a new correction coefficient FAF.

On the other hand, if the lean rich flag based on the voltage value represented by the signal S3 is "0" in step S4-2, it is determined that the air-fuel ratio is lean, and the procedure is executed to make the air-fuel ratio rich. do. That is, step S1
0, set the flag CAFR to 0 and perform step S11.
Then, it is determined whether the flag CAFL is zero. When shifting to the lean side for the first time, the flag CAFL is zero, so proceed to step S12, add a predetermined value α2 to the correction coefficient FAF, and use the result as a new correction coefficient.
FAF. In step S13, the flag
Let CAFL be 1. Therefore, if it is determined to be dilute twice or more in succession in step S4-2, a negative determination will always be made in step S11 passed from the second time onwards.
In step S14, a predetermined value β2 is added to the correction coefficient FAF, the result is set as a new correction coefficient FAF, and the FAF calculation is ended.

Note that α1, α2, β1, and β2 in steps S7, S8, S12, and S14 are predetermined values.

The feedback correction coefficient FAF determined by this calculation means is shown in FIG. 8 together with the lean rich flag of the air-fuel ratio A/F, which is expressed by filtering the voltage value represented by the air-fuel ratio signal S3. Referring to this figure, when the air-fuel ratio switches from lean to rich or from rich to lean, the correction coefficient
FAF is skipped by α1 or α2, and if it remains lean, a predetermined number β2 is successively added, and if it remains rich, a predetermined number β1 is successively subtracted.

Learning correction coefficient determined by the control method of the present invention
FG can be expressed by the following equation.

FG = (1 + FHAC + DFC / Q ... (2) where, FHAC = learning correction coefficient for altitude compensation DFC = learning correction coefficient for afrometer blockage compensation Q = intake air amount The learning correction coefficient FG is shown in Figures 1 and 9. Calculated according to the routine shown in the figure.

The learning control routine 1 shown in FIG. 1 is started every time the above-mentioned correction coefficient FAF is skipped, and in step S21, the learning control routine 1 shown in FIG.
and the previous correction coefficient FAFO, that is, the arithmetic average value FAFAV1 of the two old and new values. Proceeding to step S22, it is determined whether the average value FAFAV1 is greater than or equal to 1, and if it is less than 1, in step S23, the altitude compensation learning amount GKF is set to "-0.002", that is, the compensation learning amount GKD is set to "-0.001". ”. Average value
If FAFAV1 is 1 or more, in step S24, the altitude compensation learning amount GKF is set to "0.002", that is, the compensation learning amount GKD is set to "0.001".

In step S25, it is determined whether the conditions are such that evaporation occurs (conditions where evaporated fuel can be sucked). For example, the determination can be made based on whether the intake air amount Q is 30 m ³ /h or more, or whether the throttle valve is opened by 2 to 5 degrees or more from the closed position. That is, if the throttle valve is completely engaged with the evaporative port (not shown) provided in the intake pipe, evaporated fuel accumulated in the canister will be sucked in, so such a state is determined. If a positive determination is made, step S26
Next, it is determined whether the above-mentioned average value FAFAV1 is set to "1" when the engine is started and is increased or decreased under predetermined conditions, that is, it is greater than or equal to the compensation learning judgment value FAFAV2. Sometimes, "0.002" is added to the judgment value FAFAV2 in step S27, and the average value FAFAV1 becomes the judgment value.
When it is smaller than FAFAV2, 0.002 is subtracted from the determination value FAFAV2 in step S28.

When a negative determination is made in step S25, or when step S27 and step S28 are completed, step S
Proceed to step 29. In step S29, it is determined whether learning conditions are satisfied. It is an essential condition that the air-fuel ratio is under feedback control, and in addition, the learning condition is satisfied, for example, when the engine cooling water temperature is 80° C. or higher. If step S29 is affirmatively determined, the process proceeds to step S30, and the correction coefficient is
It is determined whether the count value of the counter CSK that counts the number of skips in FAF is 5 or more. If an affirmative determination is made in step S30, a learning control routine 2 shown in FIG. 9 is executed in step S31. Then, in step S32, the counter CSK is reset to "0".

When a negative determination is made in step S30 or when step S32 is completed, the process proceeds to step S33, where the counter CSK is incremented by +1, and in step S34, the latest correction coefficient FAF is set as the previous correction coefficient.
This series of routines ends as FAFO.

Next, the learning control routine in step S31 will be explained with reference to FIG.

When this routine is started, it is determined in step S41 whether the throttle valve 18 is fully closed based on the idle signal S7, and if an affirmative determination is made, step S41 is performed.
Proceed to step 42. If a negative determination is made, the process advances to step S47. In step S42, it is determined whether the vehicle speed SPD is zero or not, and if an affirmative determination is made, the process proceeds to step S43, and if a negative determination is made, the process proceeds to step S47. In step S43, it is determined whether the average value FAFAV1 is 1.0 or more. This judgment determines whether the correction coefficient FHAC has been learned to make the air-fuel ratio rich or lean. If the judgment is affirmative, the air-fuel ratio is set to the lean side. If the judgment is negative, it means that the air-fuel ratio has been learned to be on the rich side. If an affirmative determination is made in step S43, the process proceeds to step S44, and if a negative determination is made, the process proceeds to step S45. In step S44, the correction coefficient FHAC guard reference value FHACI is determined. If an affirmative determination is made in step S44 and if an affirmative determination is made in step S45, step S4
Proceed to step 6.

In step S46, the latest data of the correction coefficient FHAC and the guard reference value FHACI is used to execute the calculation 3×FHAC+FHACI/4, and the result is used as the latest guard reference value.
FHACI.

In step S47, 0.03 is subtracted from the latest guard reference value FHACI obtained in step S46, the result is stored in the A register, and the next step S4
In step 8, the learning amount set in step S23 or S24 of the routine in Figure 1 is set to the correction coefficient FHAC.
Add GKF to obtain the latest correction coefficient FHAC.
Next, in step S49, the correction coefficient
It is determined whether FHAC is greater than or equal to the value in the A register, and if a negative determination is made, the process proceeds to step S50, and if an affirmative determination is made, the process proceeds to step S51. That is, if the correction coefficient FHAC is smaller than (guard reference value FHACI - 0.03), the correction coefficient is
Let FHAC be (guard reference value FHACI - 0.03).

In step S51, it is determined whether the correction coefficient FHAC is greater than or equal to -0.20 and less than or equal to 0.10, and if it is not within that range, in step S52, the correction coefficient is
Guard FHAC at -0.20 or 0.10, that is, end this routine without learning the compensation learning correction coefficient DFC. In step S51, if the correction coefficient FHAC is within the range, step S5
Proceed to step 3. In step S53, it is determined whether the throttle valve 18 is fully closed, and if it is fully closed,
In step S54, the judgment value FAFAV2 is 0.98.
With the above, it is determined whether the value is 1.02 or less. If it is within that range, in step S55, the compensation correction coefficient DFC is set in step S55 of the routine of FIG.
Learning amount set in 23 or S24
Add GKD. Then, in step S56,
Add 0.002 to the judgment value FAFAV2 and end this series of routines.

FHAC and DFC learned in this way
Based on this, the learning correction coefficient FG is determined from the second equation, and is reflected in the first equation for determining the fuel injection time τ.

In this example, the compensation learning correction coefficient
A judgment value FAFAV2 related to the average value FAFAV1 is used to learn the DFC. This judgment value
FAFAV2 gradually approaches the average value FAFAV1 when the conditions for sucking evaporated fuel from the canister into the intake pipe (evaporation generation conditions) are satisfied, and therefore, under evaporation generation, the judgment value is
It is made to be smaller than 1.0. It may also be smaller than 1.0 while climbing at high altitudes. Then, this judgment value FAFAV2 is within a predetermined range, for example, within a range of 0.98 or more and 1.02 or less, under driving conditions that are not affected by the occurrence of evaporation, or when not climbing a hill at a high altitude, that is, the learning correction coefficient for compensation.
DFC learning is now performed. Therefore, when learning the correction coefficient DFC, it is possible to prevent the influence of high altitudes and evaporated fuel. In particular, step S2 in FIG.
Judgment 5 can be made depending on whether the idle signal is off or not, but by making a judgment as to whether or not the evaporation generation condition is present as in this example, the following effects can be obtained. It will be done.

Assume that the engine is now being operated in the state shown in FIG. 10A. Here, the condition under which the idle signal is on and no evaporation occurs is when the throttle valve is slightly opened to the extent that it does not touch the evaporation port and the load is light. In such a case, if step S25 in FIG. 1 is determined based on whether or not the idle signal is off, the following problem arises. 10th
Referring to Figures B and C, after time t1, the feedback correction coefficient FAF becomes smaller due to the influence of the evaporator, and its average value FAFAV1 becomes smaller.
The judgment value FAFAV2 also becomes smaller, and the correction coefficient FHAC is changed each time step S48 in FIG.
becomes smaller. FHAC does not become smaller than the lower limit value determined in step S47 of FIG. 9, for example, 0.97. Therefore, even though evaporated fuel is being sucked into the intake pipe between time points t2 and t3, the correction coefficient is
Since FHAC maintains 0.97, the feedback correction coefficient FAF becomes even smaller. As a result, the judgment value FAFAV2 also becomes smaller.

On the other hand, after time t3, evaporated fuel is no longer sucked into the intake pipe, and FAF has already become a small value.
And the air-fuel ratio becomes lean due to the influence of FHAC,
As a result, the feedback correction coefficient FAF becomes large, so the determination value FAFAV2 also becomes large, and the correction coefficient FHAC becomes large. in this way,
FAFAV2 and FHAC increase between time points t3 and t4, but FHAC is updated every five times FAF is skipped. Judgment value
FAFAV2 is getting bigger and faster. now,
Assume that the idle signal is turned on (the throttle valve is fully closed) at time t4 as shown in FIG. It has not yet returned to 1.0 and is a small value influenced by Evapo.
However, after time t4, steps S51, S53, and S54 in FIG. 9 are answered in the affirmative, and the learning correction coefficient DFC for compensation is learned in step S55, which has nothing to do with the original clogging of the airflow meter. It will show the value. Also,
By learning DFC, learning correction coefficient for altitude compensation
The FHAC value becomes smaller by the learning amount of DFC, and as a result, the guard reference value FHACI, which is the reference for the lower limit value of FHAC, becomes undesirably small.

Therefore, in step S25 of FIG. 1, the 10th
After time t3 in the figure, steps S26 and S2 in FIG.
7. Prevented the process from passing through S28. As a result, as can be seen from FIG. 10E, time points t3 to t
5, the judgment value FAFAV2 is not updated, therefore,
The determination of 1.020.98 in step S54 of FIG. 9 is denied, and as a result, erroneous learning of the correction coefficient DFC is prevented.

In addition, in this embodiment, steps S41 to S in FIG.
Since the lower limit of the correction coefficient FHAC is determined by 50, the correction coefficient FHAC does not become too small due to the influence of the evaporative pump when driving on flat ground, and the correction coefficient FHAC can be adjusted to the normal value by learning when driving without the evaporative pump. quickly returns to the value of , and therefore,
The influence of evaporated fuel on altitude compensation can be minimized. In addition, when updating the guard standard value FHACI, an average value indicating the so-called pace air-fuel ratio will be added.
Determine whether FAFAV1 is less than or equal to 1,
According to the result of that determination, that is, when the current pace air-fuel ratio is on the lean side, FHACFHACI
When the pace air-fuel ratio is on the rich side,
Guard reference value only when FHACFHACI
Since the FHACI is updated, the correction coefficient can be adjusted even if the evaporation occurs under certain driving conditions, such as when driving in LA4 mode in the United States.
Since the lower limit value of FHAC is regulated according to the correct guard reference value FHACI, the correction coefficient FHAC is also correctly learned. This point will be explained in detail with reference to FIGS. 11A to 11E.

Assume that the engine is now being operated in the state shown in FIG. 11A. If the guard reference value FHACI is constantly updated when the idle signal is on, the following problems arise. Figure 11B
As shown in , the guard reference value FHACI is at time t
1 to t2, the idle signal is off, that is, the throttle valve is open, so it is not updated and maintains the initial value of 1.0. On the other hand, as shown in FIG. 11C, the correction coefficient FHAC is affected by the evaporation from point t1. In other words, because the average value FAFAV1 indicating the pace air-fuel ratio has become smaller than 1.0 due to the occurrence of evaporation, the correction coefficient FHAC becomes smaller every time the learning routine is performed, but the first lower limit value (guard reference value) FHACI−0.03),
It does not become smaller than the lower limit value of 0.97. Furthermore, between time points t2 and t3, the idle signal is turned on, that is, the throttle valve is fully closed, and evaporation generation is stopped. Therefore, the average value FAFAV1 becomes larger than 1.0 due to the correction coefficient FHAC which has been learned to a small value under the occurrence of evaporation, 0.97 in Fig. 11C, and as a result, the correction coefficient increases each time the learning routine is performed. Become. On the other hand, from time t2 to t3, the guard reference value FHACI can be updated,
gradually decreases according to the value of the correction coefficient FHAC,
When the correction coefficient FHAC and the guard reference value FHACI approach each other, the guard reference value FHACI gradually increases, but if the idle signal turns off at time t3 before the guard reference value FHACI returns to 1.0, the value at time t3 0.99 becomes the guard reference value FHACI at time t3 to t4. Therefore, time t
3 to t4, the lower limit of the correction coefficient FHAC is the second
The lower limit of is 0.96. If such operating conditions continue, the lower limit value will become even smaller, and the correction coefficient FHAC will become too small due to the occurrence of evaporation, making it difficult to perform the original altitude compensation.

As in the embodiment of the present invention, the guard reference value FHACI
By taking into consideration the determination in step S43 in FIG. 9 when updating the FHAC, even if the throttle valve is fully closed, the FHAC
The guard reference value FHACI is not updated unless it becomes FHACI (see Figure 11D), and therefore the first
As shown in FIG. 1E, the lower limit value of the correction coefficient FHAC remains at 0.97 even between time points t3 and t4. Therefore, even if such an operating condition continues, the lower limit value will not become extremely smaller than 0.97, and the original altitude compensation can be reliably performed without being affected by the evaporation system.

〔Effect of the invention〕

According to the present invention, the correction coefficient DFC for compensating for clogging of the air flow meter due to changes over time is:
At least, the information is updated and learned only when the average value FAFAV1 of the feedback correction coefficient FAF and the related judgment value FAFAV2 are within a predetermined range. only under evaporative conditions)
Since the judgment value FAFAV2 was made to be close to the average value FAFAV1, the judgment value
FAFAV2 becomes a small value and becomes a value outside the above range, and therefore updating of the correction coefficient DFC is prohibited.
Additionally, the condition for evaporation generation is that the throttle valve must be opened at least to the extent that it touches the evaporation port of the intake pipe (the port through which evaporated fuel from the canister is sucked into the intake pipe); Even if you climb a hill to a high altitude with the airflow meter opened more than a certain degree, the judgment value FAFAV2 will deviate from the above judgment value, so in this case as well, updating the correction coefficient DFC for compensating for clogging of the airflow meter is prohibited. Therefore, the correction coefficient DFC for compensating for clogging of the air flow meter is not affected by the evaporator or high altitude, and learning is performed in accordance with the original purpose.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing a learning control routine that is an embodiment of the present invention, FIG. 2 is a chart showing the influence of the air-fuel ratio due to evaporated fuel, FIG. 3 is a diagram showing the influence of the air-fuel ratio due to high altitude, and FIG. FIG. 5 is a diagram showing the influence of air-fuel ratio due to blockage of intake air amount, FIG. 5 is a configuration diagram showing an example of an internal combustion engine to which the method of the present invention is applied,
FIG. 6 is a block diagram showing a detailed example of the control circuit, FIG. 7 is a flowchart showing an example of the feedback correction coefficient, and FIG. 8 is a time chart showing the flag and correction coefficient FAF according to the air-fuel ratio signal S3. FIG. 9 is a flowchart showing an example of the learning control routine 2 of the routine in FIG.
Figure 0A to F are the idle signal and judgment value, respectively.
FAFAV2, correction coefficient FHAC, correction coefficient DFC time chart, Figure 11 A to E are the idle signal, guard reference value FHACI, and correction coefficient, respectively.
This is FHAC's time chart. 10... Engine body, 18... Throttle valve, 20...
Air flow meter, 22...control circuit, 34, 36...
Crank angle sensor, 40...Idle switch, 4
2... _O2 sensor.

Claims (1)

[Claims] 1 Calculates the basic fuel injection time TP based on the intake air amount Q and engine speed NE, and performs feedback according to the measured air-fuel ratio so that the air-fuel ratio becomes the stoichiometric air-fuel ratio under predetermined feedback conditions. The feedback correction coefficient FAF is calculated by a predetermined number in response to the measured air-fuel ratio changing from rich to lean or from lean to rich.
Calculate the arithmetic average value FAFAV1 of the old and new values immediately before or after the FAF skips, determine the conditions for the vaporized fuel in the canister to be sucked into the intake pipe, and when the above conditions are satisfied, Compare the judgment value FAFAV2 for determining the learning condition of the compensation learning correction coefficient DFC with the average value FAFAV1, and if the judgment value FAFAV2 is larger than the average value FAFAV1, reduce the judgment value FAFAV2, and if smaller, increase the judgment value FAFAV2. However, when the engine is idling and the judgment value FAFAV2 is within a predetermined range indicating that the fuel vapor in the canister is not affected by being sucked into the intake pipe or when the engine is not climbing a hill at a high altitude, that is. The value of the compensation learning correction coefficient DFC is updated and learned, and the final fuel injection time τ is determined based on at least the basic fuel injection time TP, the feedback correction coefficient FAF, and the compensation learning correction coefficient DFC. An air-fuel ratio control method characterized by: