JPH0783099A - Air-fuel ratio controller of engine - Google Patents

Abstract

PURPOSE:To make fuel supply optimum by utilizing a correlation method using a simulated irregular signal so as to calculate a fuel corrective quantity of an optional cylinder. CONSTITUTION:An M series signal epsilon and air-fuel ratio sensor output (y) based on the input of the signal epsilon are data-inputted for every constant sampling cycle (S11), and relative correlation function phiepsilony (alpha) from epsilon and (y) per sampling cycle is calculated (S13), and impulse response is calculated (S14), stepping responsive is calculated (S15), and then a corrective quantity relating to a basic injection pulse width Tp is calculated from the step response (S16).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an engine air-fuel ratio control device, and more particularly to a device for controlling an air-fuel ratio for each cylinder.

[0002]

2. Description of the Related Art There is an air-fuel ratio feedback control device for properly controlling the fuel supply amount of an engine. According to this device, the fuel injection amount is determined based on the output from an air-fuel ratio sensor provided in an exhaust system. The air-fuel ratio is constantly kept constant by correcting (see Automobile Technology Handbook (Company) Automotive Engineering Society, Second Volume, page 162).

[0003]

However, in such an air-fuel ratio feedback control device, since the injection amount of all cylinders is the same, the difference in the characteristics of the injectors mounted in each cylinder and the cam lift of each cylinder. The difference in the air-fuel ratio between the cylinders caused by the difference in the air amount caused by the above was unavoidable. For this reason, the air-fuel ratio during the air-fuel ratio feedback control changes slightly, resulting in a decrease in the conversion efficiency of the three-way catalyst and the like, and in any case, it was difficult to obtain the best conversion performance.

By the way, in a general control system including such an air-fuel ratio feedback control device, if the relationship generated between the input signal and the output signal, that is, the transfer function is obtained for each cylinder, the output You can fully predict the changes,
By correcting the input based on the transfer function, it is possible to always realize control with small fluctuation. The input signal X
(S) and the output signal Y (S), the transfer function G
(S) is obtained as G (S) = Y (S) / X (S).

However, the fuel injection amount signal, which is an input signal in the case of air-fuel ratio control, is greatly affected by fluctuations in the output due to intake air pulsation of the air flow meter, so that the output signal, such as generated torque and exhaust composition, is affected. May not be an accurate value, and thus the transfer function G
(S) cannot be obtained accurately. Further, even if it is required, it is almost meaningless to apply it to each of the production engines from the viewpoint of enhancing the individual controllability due to the above-mentioned individual variations.

The present invention has been made by paying attention to such a conventional problem, and estimates a unique dynamic characteristic of a control system of an arbitrary cylinder of the engine by utilizing a correlation method using a pseudo-random signal. However, it is an object of the present invention to provide an apparatus that optimizes the fuel supply by calculating the fuel correction amount of an arbitrary cylinder based on this.

[0007]

Therefore, the present invention is based on FIG.
As shown in, the engine load (for example, intake air amount Q
a) and sensors 21 and 22 for respectively detecting the rotation speed Ne
In the air-fuel ratio control device for an engine, which is provided with a means 23 for calculating the basic injection amount Tp of fuel from the detection values of these sensors and a means 41 for outputting a signal of the basic injection amount Tp to the fuel supply device 42, Cylinder selecting means 40 for selecting a cylinder, and a periodic minute pseudo irregular signal (for example, M
A device 26 for generating a sequence signal ε), means 27 for superimposing this pseudo-random signal ε on the signal of the basic injection quantity,
The sensor 28 that detects the air-fuel ratio of the burned gas when the pseudo-random signal ε is superimposed, and the corresponding air-fuel ratio sensor output y when the pseudo-random signal ε is also superimposed.
And the pseudo-random signal ε, the cross-correlation function φεy of both signals
And means 30 for calculating the impulse response g (α) from the cross-correlation function φεy, and means 31 for obtaining the step response r (α _L ) by integrating the impulse response g (α). A means 32 for setting a fuel increase / decrease correction amount HOS corresponding to the obtained step response r (α _L ), a means 33 for adding the increase / decrease correction amount HOS to the basic injection amount Tp at the transition, And a means 24 for outputting to the supply device 252.

[0008]

The distribution of the air-fuel ratio among the cylinders is caused by the difference in the injector characteristic and the intake valve lift amount of each cylinder. On the other hand, according to the present invention, the air-fuel ratio distribution (A / F) for each cylinder is
The increase / decrease correction amount HOSn of each cylinder that eliminates (distribution) is added to the basic injection amount. However, n is a subscript indicating a cylinder.

In this case, since the correlation method in which the pseudo-random signal is superimposed is used, the gain of the step response r (α _L ), that is, the A / F distribution of each cylinder can be accurately obtained. Therefore, the increase / decrease correction amount HO calculated in consideration of this A / F distribution
Since S also gives an accurate value, accurate increase / decrease correction of each cylinder is performed. Further, the accuracy is further improved under the operating condition where the exhaust gas of each cylinder is well mixed.

[0010]

FIG. 2 is a system diagram of an embodiment. In the figure, the intake air amount Qa is detected by the air flow meter 2, the engine speed Ne is detected by the crank angle sensor 3, and the air-fuel ratio in the exhaust gas is detected by the air-fuel ratio sensor 4. These signals are input to the control unit 5, and the control unit 5 optimizes the fuel injection amount of each cylinder according to the flowcharts shown in FIGS.

Reference numeral 6 is an injector (fuel supply device) for outputting a fuel injection amount signal, and 7 is an ignition coil. FIG. 3 is a routine for superimposing the M-sequence signal, which is one of the pseudo-random signals, on the fuel injection amount signal. In S1, it is determined whether the operating condition is in the steady state, and if it is in the steady state, S2 is executed.
Proceed to.

In S2, cylinder selection is performed based on the signal from the crank angle sensor 3 and the like. S3 is a portion which fulfills the function of the signal superimposing means 27 of FIG. 1, and here superimposes the M-sequence signal ε (t) on the fuel injection pulse width (described later) TiB calculated by the conventional device. That is, the fuel injection pulse width Ti in this case is Tin = TiB + ε (t). When this Tin is given to the injector 6 of the n-th cylinder (n is a subscript indicating a cylinder), the injector 6 injects fuel toward the intake port of the engine during this Ti.

As shown in the upper part of FIG. 6, the M-sequence signal ε (t) has an amplitude a, a minimum pulse width Δ, and one cycle NΔ (N is the maximum sequence, which is 15 in the embodiment, but is 7, 31). Can also be used). Therefore, the autocorrelation function φεε (τ) is also a periodic function and becomes a triangular pulse train having a narrow width as shown in the lower part of FIG. This M-sequence signal is a very small signal, and even if it is superimposed on the injection amount signal, the fluctuation of the engine rotation is small,
It does not impair the driving sensation of the driver.

The signal is not limited to the M-series signal, and an L-series signal or a twin prime number sequence signal may be used as long as it is a pseudo pseudo-random signal. FIG. 4 is a routine for obtaining the fuel increase / decrease correction amount HOSn for each cylinder with respect to the basic injection pulse width Tp. Steps S11 to S14 are portions for obtaining an impulse response by a correlation method using an M-sequence signal, and this correlation method itself is known.

First, this method will be described using mathematical expressions. Input signal to engine control system (plant) x
(T), when the output signal based on this input is the air-fuel ratio sensor output y (t), these are expressed by the following equations (1) and (2). x (T) = ε (t) + μ (t) ... (1) y (T) = ρ (t) + ξ (t) ... (2) where ρ (t) corresponds to the M-sequence signal ε (t). The output components, μ (t) and ξ (t), are DC components. The reason is that μ (t) is an original input of the engine control system and can be regarded as a DC component because its period is longer than the period of the M-sequence signal.

If the amplitude a of the M-series signal ε (t) is sufficiently small, the combustion efficiency characteristic of the engine (output torque characteristic with respect to the fuel amount) within that amplitude can be regarded as linear, so that the M-series signal ε ( The relationship between t) and the output component ρ (t) is expressed by the equations (3) to (5) using the impulse response g (τ).

[0017]

[Equation 1]

Further, the cross-correlation function φερ (α) between ε (t) and ρ (t) is expressed by the equation (6).

[0019]

[Equation 2]

Where φεε (α) is the autocorrelation function of the M-sequence signal,

[0021]

[Equation 3]

Is given by On the other hand, M-sequence signal ε (t)
Contains all frequency components, its power spectral density function φεε (ω) is constant, so φεε
(Ω) = φεε (0). As a result, the autocorrelation function φεε (α−τ) in the equation (6) is expressed by the equation (8) using the delta function δ.

Φεε (α−τ) = Φεε (0) · δ (α−τ) (8) Therefore, the cross-correlation function φερ shown in the equation (6)
(Α) is transformed as follows.

[0024]

[Equation 4]

As is clear from the above equation (9), the impulse response g (α) is the cross-correlation function φ of ε (t) and ρ (t).
It is given by equation (10) using ερ (α). g (α) = φερ (α) / Φεε (0) ... (10) Here, Φεε (0) corresponds to the integral value of the autocorrelation function φεε, and Φεε (0) = (N + 1) Δ · α ² / N = Z (constant) is given by (11).

The cross-correlation function φερ (α) is given by the following equation from the equation (2).

[0027]

[Equation 5]

Therefore, g (α) = {φεy (α) -φεξ (α)} / Z ... (13) Here, the second term φεξ (α) in the equation (13) is M
It is a cross-correlation function of the series signal ε (t) and the DC component ξ (t) of the signal. The first term φεy (α) is the M-sequence signal ε
It is a cross-correlation function between (t) and the output y (t). y
(T) is a fluctuation component due to the influence of the M-sequence signal ε (t),
Although it is composed of a DC component, it is difficult to detect the component separately, and the cross correlation function φεy shown in the following equation (13) ′ is directly obtained.

[0029]

[Equation 6]

Here, the value of φεξ (α) is the value of α by ε
If it is made sufficiently large until the influence of (t) disappears, φεy
It matches the value of (α). Therefore, φεξ (α)
It can be approximated by the average value g (α) in the sections α ₁ and α ₂ of εy (α).

[0031]

[Equation 7]

However, the values of α at the time when the impulse response g (α) is sufficiently settled are used as α ₁ and α ₂ indicating the integration range. For α ₂ −α ₁ , select a value close to N · Δ. The above is the procedure for obtaining the impulse response. In actual control, the step response is easier to control than the impulse response. In this case, since the step response is the integral of the impulse response, if the impulse response g (α) obtained by the equation (14) is integrated in the sections α _{S to} α _L , the time α
Step response in _L r (alpha _L) is given by (15).

[0033]

[Equation 8]

However, α _S is the integration start time (close to zero) in consideration of the deviation of the rising edge of the impulse response due to the pseudo whiteness of the M-sequence signal. α _L is the end time of the integration section when the impulse response is integrated, and is set in advance according to the characteristics of the impulse response. Note that r (α _L )
Is normalized by Z shown in equation (13), and therefore corresponds to the step response when a unit input is given.

This completes the logical explanation. In practice, the output of the air-fuel ratio sensor, which is an output signal, is sampled at regular intervals, and digital signal processing is performed in the control unit. Therefore, the cross-correlation function φεy
The integral value required to obtain (α) and each response g (α), r (α _L ) is replaced by the integrated value. In other words, the actual control system is composed of discrete value systems.

Returning to FIG. 4, in S11, the M-series signal ε and the air-fuel ratio sensor output y based on this input are data-input and stored at constant sampling intervals. In S12, it is determined whether or not the data input for one period of the M-sequence signal ε and y is completed, and if completed, the process proceeds to S13. S1
Reference numeral 3 denotes a portion which fulfills the function of the cross-correlation function calculation means 29 of FIG. 1, and here, the cross-correlation function φεy (α) of ε and y for each sampling period is calculated by the equation (13) ′.

S14 is the impulse response calculation means 3 of FIG.
In this part, the impulse response g (α) is obtained by the equation (14). S15 is a portion that fulfills the function of the step response calculation means 31 of FIG. 1, and here, the impulse response g (α) is used to obtain the step response r (α _L ) from the equation (15).

In this case, r (α _L ) represents the value at time α _L after the fuel is changed stepwise,
As shown in the upper part of FIG. 7, α _L1 , α
If distinguished from _L2 , α _L3 , ..., Step responses r (α _L1 ), r (α _L2 ), r (α _L3 ), ...
Is the value shown in the upper part of FIG. Note that rn (α _L ) is n
No. cylinder is shown.

FIG. 7 shows that when the fuel in the specific cylinder is increased stepwise by 1, the fuel flowing into the cylinder flows in with a response delay as shown by the curve. Note that a 4-cylinder engine is shown, and one cylinder change results in a 1/4 change. S16 is a portion which fulfills the function of the increase / decrease correction amount setting means 32 of FIG. 1, and here, the correction amount HOS for the basic injection pulse width Tp is obtained from the step response r (α _L ).

In the upper part of FIG. 7, since 1-rn (α _L8 ) is the fuel that is insufficient as the distribution difference between the cylinders after convergence,
This 1-rn (α _L8 ) must be increased as a correction amount. Here, α _L8 is defined as the time when the step response converges sufficiently. Since the step response r (α _L ) shown in the upper part of FIG. 7 is for the case where the fuel amount is increased by 1, actually 1-r (α _L ) depends on the increment step width of Tp. The correction amount is obtained by multiplying by a constant.

In S17, the calculated correction amount HOS is stored. FIG. 5 is a routine for calculating the fuel injection pulse width. At S21, the intake air intake amount Qa and the engine speed Ne are read, and the routine proceeds to S22. S22 is a portion that fulfills the function of the basic injection amount calculation means 23 in FIG. 1, and here, the basic injection pulse width Tp (= K · Qa / N, where K is a constant giving the basic air-fuel ratio) is obtained by referring to a map or the like. .
In this case, Tp gives a value for the steady state.

S24 is a portion which fulfills the function of the adding means 33 of FIG. 1, and here the fuel injection pulse width Ti is calculated by the following equation (17). Tin = (Tp + HOSn) * Co + Ts (17) where Co is the sum of 1 and the water temperature increase correction coefficient, and Ts is the invalid pulse width. 3 to 5 are routines given to the CPU, the block diagram of FIG.

Then, when the fuel injection pulse width Tin is set in the output register and a predetermined fuel injection timing in synchronization with engine rotation is reached, a drive pulse signal having a calculated pulse width of the fuel injection pulse width Tin is injected. 6, fuel injection is performed. here,
The operation according to the embodiment will be described.

A / F difference between the cylinders occurs due to variations in the injectors and the cam lift amount of the cylinders, resulting in a decrease in conversion efficiency. On the other hand, according to the embodiment,
The A / F difference of each cylinder is introduced as a correction amount HOSn, and this correction amount HOSn is added to Tp of each cylinder. Even if the M-sequence signal is superposed in order to optimize the fuel injection amount, this signal is very small, and since it is performed during steady operation, the drivability during superposition is not impaired.

Finally, in the embodiment, the case of step increase has been described, but it goes without saying that the same applies to the case of step decrease.

[0046]

As described above, according to the present invention,
The step response is obtained by using the correlation method with superimposing a pseudo-random signal, the fuel correction amount is calculated in accordance with the convergence value of this step response, and this correction amount is added to the basic injection amount. Since the fuel injection amount for each cylinder is optimized, the fuel can be supplied to each cylinder without excess or deficiency, and the fuel supply can be optimized, for example, the conversion efficiency of the three-way catalyst can be improved. There is an effect.

[Brief description of drawings]

FIG. 1 is a claim correspondence diagram showing a configuration of the present invention

FIG. 2 is a system diagram showing an embodiment of the present invention.

FIG. 3 is a flow chart for explaining the sensor operation of the above embodiment.

FIG. 4 is a flowchart for explaining the sensor operation of the above embodiment.

FIG. 5 is a flowchart for explaining the sensor operation of the above embodiment.

FIG. 6 is a waveform diagram showing an M-sequence signal ε (t) and an autocorrelation function φεε (τ) of this signal.

FIG. 7 is a waveform diagram for explaining the operation of the embodiment.

FIG. 8 is a block diagram showing another embodiment of the present invention.

[Explanation of symbols]

 2 Air flow meter (engine load sensor) 3 Crank angle sensor (engine speed sensor) 4 Air-fuel ratio sensor 5 Control unit 6 Injector (fuel supply device) 21 Engine load sensor 22 Engine speed sensor 23 Basic injection amount calculation means 24 Output means 25 Fuel supply device 26 Pseudo irregular output generator 27 Signal superimposing means 28 Air-fuel ratio sensor 29 Cross-correlation function calculating means 30 Impulse response calculating means 31 Step response calculating means 32 Increase / decrease correction amount setting means 33 Addition means

Claims (2)

[Claims] 1. A sensor for detecting a load and a rotational speed of an engine, a means for calculating a basic injection amount of fuel from detection values of these sensors, and a means for outputting a signal of the basic injection amount to a fuel supply device. In an air-fuel ratio control device for an engine, the device for generating a periodic minute pseudo irregular signal, means for superimposing the pseudo irregular signal on the signal of the basic injection amount of an arbitrary cylinder, and the pseudo irregular signal A sensor that detects the air-fuel ratio of the burned gas when the signal is superimposed, and a cross-correlation function of both signals from the corresponding air-fuel ratio sensor output and the pseudo-irregular signal when the pseudo-irregular signal is also superimposed. , A means for calculating an impulse response from the cross-correlation, a means for obtaining a step response by integrating the impulse response, and a means for calculating the step response Air-fuel ratio control of an engine, comprising means for setting an increase / decrease correction amount of fuel in an arbitrary cylinder in accordance with the response of the cylinder and means for adding the increase / decrease correction amount to a basic injection amount of the arbitrary cylinder. apparatus. 2. The air-fuel ratio control apparatus for an engine according to claim 1, further comprising means for sequentially switching an arbitrary cylinder, calculating an increase / decrease correction amount for all cylinders, and adding the calculated increase / decrease correction amount to a basic injection amount.