JPH06173743A - Air-fuel ratio learning control method for internal combustion engine - Google Patents

Abstract

PURPOSE:To perform correction with a small memory capacity and at high speed by determining a common learning value in the whole area by defining the deterioration characteristic of emission parts with a weighting factor. CONSTITUTION:A weighting factor by which the deterioration characteristic of an air flow sensor is defined for the whole area is stored in a weighting factor deciding means 102 beforehand. The weighting factor deciding means 102 decides the weighting factor KA corresponding to the intake air quantity detected by a running state detecting means 101. A steady state learning means 103 determines learning correction quantity of this time by judging a specified steady state, sampling the feedback correction quantity of the air fuel ratio decided according to the output of an O2 sensor, and dividing this feedback correction quantity by the weighting factor KA. A learning memory means 104 stores the learning correction value of this time reflected on the common learning value already stored. This learning value corrects fuel injection quantity as an air fuel ratio correction factor by means of fuel injection quantity correcting means 106 by multiplying the weighting factor KA by a learning correction quantity calculating means 105.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an air-fuel ratio learning control method for an internal combustion engine capable of obtaining an optimum air-fuel ratio in all operating regions.

[0002]

2. Description of the Related Art Conventionally, in a system in which a fuel injection amount is corrected by the output of an O ₂ sensor and the correction amount is stored as a learning value, the correction amount is adjusted in accordance with the operating state of the engine, such as rotation speed, load, etc. It was divided into subsections and stored. However, such a learning method has a drawback that the learning speed becomes slow because it is necessary to stay in a steady state in each region. With respect to such a problem, Japanese Patent Laid-Open No. 63-259136 discloses a learning method for estimating an unlearned region according to a predetermined deterioration characteristic of an air flow sensor.

[0003]

However, since the learning result in a certain reference region is used even by such a method, there is a problem that a sufficient learning speed cannot be obtained and a large memory capacity is required. is doing. In view of the above problems, the present invention defines a deterioration characteristic of an emission part by a weighting coefficient to obtain a single learning value common to all areas, thereby enabling correction with a small memory capacity and at a high speed. An object of the present invention is to provide an air-fuel ratio learning control method for an engine.

[0004]

In order to achieve the above object, the present invention defines the deterioration characteristic of the air flow sensor as a weighting coefficient and divides the deviation amount of the air-fuel ratio in each region by the weighting coefficient. The values are averaged to obtain one common learning value, and the actual fuel injection amount is corrected using a value obtained by multiplying the learning value by a weighting coefficient corresponding to the operating region as an air-fuel ratio correction coefficient.

[0005]

By the above means, the learning correction can be performed in the entire area without dividing the learning area into the small areas. Therefore, an air-fuel ratio learning control method with a small memory capacity and a high learning speed can be realized.

[0006]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An air-fuel ratio learning control method for an internal combustion engine of the present invention
An example will be described with reference to the drawings. In this example,
The deterioration characteristic of the airflow is adopted as the deterioration characteristic of the emission parts, and the operating region is divided by the intake air amount to define the weighting coefficient. FIG. 2 is a schematic view of a main part of an internal combustion engine to which the air-fuel ratio learning control method of this example is applied. 1 is an engine body, 2 is an air cleaner, 3 is an air flow sensor,
4 is an intake pipe, 5 is a throttle valve, 6 is a throttle chamber, 7 is an intake manifold, 8 is an injector, 9 is an exhaust pipe, 10 is an O ₂ sensor, 11 is a three-way catalytic converter, 12 is a water temperature sensor, and 13 is a crank. It is a shaft.

Reference numeral 16 is an ECU (electronic control unit) to which the following is input. That is, the intake air amount detected by the air flow sensor 3, the opening degree of the throttle valve 5 detected by the throttle sensor 17,
Engine cooling water temperature detected by the water temperature sensor 12,
The engine speed detected by the rotation sensor 18 connected to the crankshaft 13 and the oxygen concentration detected by the O ₂ sensor 10 are input.

The ECU 16 determines the basic energization time of the injector in accordance with each of the input signals, and further corrects the basic energization time by the learning correction value calculated by the predetermined learning value calculating means to finally obtain the final energization time. Determine the injection amount. On the other hand, the output of the ECU 16 is connected to the injector 8 and
₂ Depending on the oxygen concentration detected by the sensor 10,
The fuel injection amount of the injector 8 is feedback-controlled by determining the overthickness.

Here, the fuel injection pulse width generally calculated in the conventional ECU 16 is as shown in the following [Equation 1]. [Equation 1] T _i = K ₁ × Q _A / N × K ₂ × α + T _V T _i ... Fuel injection pulse width, K ₁ ... Constant, Q _A ... Intake air amount, N ... Engine speed, K ₂ ... Water temperature , Correction coefficient for transient, α ... Feedback correction coefficient, T _V ... Invalid injection time.

Normally, the feedback correction coefficient α used in [Equation 1] is adjusted so that it can be controlled around 1.0, but with the passage of time, each sensor or injector 8 and the like deteriorates with time, and α Is controlled when the value of is deviated from 1.0. On the other hand, in order to start the O ₂ feedback, it is necessary to complete the warm-up of the O ₂ sensor 10, and α is fixed at 1.0 until then. Therefore, if the value of α deviates due to the deterioration of the system over time, the base air-fuel ratio before the start of feedback becomes a deviated value. Further, the control range of α is generally controlled to about ± 25% in terms of responsiveness in terms of control, so that feedback control cannot be performed when the air-fuel ratio deviates from this value.

Therefore, in the present invention, the deviation of the feedback correction coefficient α from 1.0 is stored in a back-up storage device and used for calculation of the fuel injection pulse width. At this time, the calculation formula is as follows. [Equation 2] T _i = K ₁ × Q _A / N × K ₂ × (α + K _L ) + T _V Here, K _L is an air-fuel ratio correction coefficient, which is stored by using [Equation 5] described later. The correction amount and the weighting coefficient are obtained as the final learning correction amount.

Next, FIG. 1 shows a control block in the ECU 16 for calculating the final learning correction amount K _L. The operating state detecting means 101 detects the intake air amount Q _A in this example. The weighting factor determining means 102 stores in advance a weighting factor that defines the deterioration characteristic of the air flow sensor 3 for the entire region of the intake air amount Q _A as shown in FIG.
The weighting factor detecting means 102 determines the weighting factor K _A corresponding to the intake air amount Q _A detected by the operating state detecting means 101.

The steady learning means 103 determines a predetermined steady state, samples the feedback correction amount of the air-fuel ratio determined according to the O ₂ sensor output, and weights the coefficient K _A determined by the weighting coefficient determining means 102. This feedback correction amount is divided by this learning correction value K
Calculate as _{LN (i)} . The learning value storage unit 104 reflects all or part of the common learning value K _LN that has already stored the learning correction value K _{LN (i)} of this time and stores it. The learning value K _LN stored in this way is multiplied by the weighting coefficient K _A determined by the weighting coefficient determination means 102 in the learning correction amount calculation means 105 to obtain the final learning correction amount, that is, the air-fuel ratio correction coefficient K _L as the fuel. The injection amount correction means 106 corrects the fuel injection amount.

The procedure for obtaining the final learning correction amount K _L described above will be described with reference to the flowcharts of FIGS. 3 and 4. First, in step 301, both the flags 1 and 2 are set to 1, and the integrated value αS of the counter and the feedback correction coefficient α is set to 0. In step 302, it is judged from the water temperature condition, the steady state judgment, etc. whether or not a predetermined learning enable condition is satisfied. If the condition is not satisfied, both flags 1 and 2 are set to 0 in step 401, and the process proceeds to step 402. If the condition is met, step 303
Up to 308, the feedback correction coefficient α is integrated as follows.

At step 303, the feedback correction coefficient α is fetched. In step 304, αS = αS + α is calculated to obtain the integrated value αS of the feedback correction coefficient α. In step 305, 1 is added to the counter. In step 306, the counter indicates the predetermined sampling number S
It is determined whether or not has reached, and if not achieved, the process returns from step 308 to step 302. As described above, the feedback correction coefficient α is integrated for each feedback cycle, and the sampling number is repeated a predetermined number of times S,
If it is determined in step 306 that the integration has been completed, the process proceeds from step 307, 308 to step 402.

In step 402, flag 2 is 1.
It is judged whether or not (whether or not the learning condition is satisfied), and N
If it is o, the process returns, and if Yes, step 309.
Then, the learning value is calculated and updated. Step 30
9, the predefined weighting factor K shown in FIG._AInhale
Air volume Q _ARead according to. In step 310, now
K as a learning value of times_{LN (i)}Is calculated as the following [Equation 3]
It [Equation 3] K_{LN (i)}= ΑS / S / K_A

At step 311, the learning value K _{LN (i)} obtained this time is updated by reflecting the learning value K _LN already stored by a predetermined learning update ratio β (β <1). [Formula 4] is calculated. [ _Equation 4] K _LN = K _LN * β + K _{LN (i)} * (1-β) This is to prevent erroneous learning and also to prevent the controllability from being impaired due to a rapid change in learning.

When the actual fuel injection amount is calculated from the learning value K _LN stored as described above, the weighting coefficient K _A is read out as in the case of calculating the learning value, and the following [Equation 5] is used. The air-fuel ratio correction coefficient K _L is calculated as the final correction amount suitable for each region. [Equation 5] K _L = K _LN * K _A Then, using this coefficient K _L , the actual fuel injection pulse width is calculated by the above [Equation 2]. With such a configuration, the learning correction can be performed in the entire area without dividing the learning area into small sections, so that an air-fuel ratio learning control method with a small memory capacity and a high learning speed can be realized.

In the embodiment described above, the weighting factor K
_A was defined only in the intake air amount direction based on the deterioration characteristics of the air flow sensor, but considering the deterioration characteristics of other emission parts, such as the injector, the engine load such as the intake pipe pressure or the intake air amount per unit rotation is also considered. It is also possible to make a two-dimensional map that is divided into regions by parameters and engine speed. In this case, the effect of using the small memory capacity of the above-described embodiment is reduced, but the merit of more precise air-fuel ratio control and high-speed learning is exhibited.

[0020]

According to the present invention, one common learning value is constantly updated in all regions of the operating state, and deterioration of emission parts can be compensated by one common learning value in all regions. Therefore, an internal combustion engine learning control method that can learn at a high speed with a small memory capacity can be obtained.

[Brief description of drawings]

FIG. 1 is a block diagram of a control block that calculates a learning correction amount according to an embodiment of the present invention.

FIG. 2 is a schematic view of a main part of an internal combustion engine to which the air-fuel ratio learning control method according to the embodiment of the present invention is applied.

FIG. 3 is a flowchart showing a procedure for obtaining a learning correction amount according to the embodiment of the present invention, part 1;

FIG. 4 is a flowchart showing a procedure for obtaining a learning correction amount according to the embodiment of the present invention, part 2;

FIG. 5 is a graph showing weighting factors used in the embodiment of the present invention.

[Explanation of symbols]

1 ... engine body 3 ... air flow sensor 8 ... injector 10 ... O ₂ sensor 16 ... ECU T _i ... fuel injection pulse width Q _A ... intake air quantity alpha ... feedback correction coefficient K _A ... weighting factor K _L ... final learning correction amount K _LN … Learning value K _{LN (i)} … Learning correction value this time

Claims (3)

[Claims] 1. Degradation characteristics of emission parts that affect air-fuel ratio control are defined by weighting factors for all operating regions, and the air-fuel ratio deviation amount of each operating region is divided by the weighting factor in the operating region. Is stored as one common learning value, and the fuel injection amount is corrected using a value obtained by multiplying the learning value by a weighting coefficient corresponding to the operating region as an air-fuel ratio correction coefficient. Air-fuel ratio learning control method. 2. The air-fuel ratio learning control of an internal combustion engine according to claim 1, wherein the operating region is divided by an intake air amount, and the weighting coefficient is a value based on a deterioration characteristic of an air flow sensor. Method. 3. The air-fuel ratio learning control method for an internal combustion engine according to claim 1, wherein the operating region is a two-dimensional map divided by an engine load parameter and an engine speed.