JPH0686839B2 - Feedback controller with learning function - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a feedback control device having a learning function applied to an electronically controlled fuel injection control device for an internal combustion engine and the like.

<Background Art> In an electronically controlled fuel injection internal combustion engine, the injection amount Ti is determined by the following equation.

Ti = Tp × COEF × α + Ts where Tp is the basic injection amount and Tp = K × Q / N.

K is a constant, Q is the intake air flow rate, and N is the engine speed.
COEF is various correction factors. α is an air-fuel ratio feedback correction coefficient for air-fuel ratio feedback control (λ control) described later. Ts is a voltage correction amount, and is for correcting a change in the injection amount of the electromagnetic fuel injection valve due to a change in the battery voltage.

For λ control, an O ₂ sensor is installed in the exhaust system to detect the actual air-fuel ratio, and the slice level is used to determine whether the air-fuel ratio is richer or thinner than the stoichiometric air-fuel ratio, and the fuel is injected to reach the stoichiometric air-fuel ratio. The amount is controlled, and therefore, the air-fuel ratio feedback correction coefficient α is defined and is changed to maintain the stoichiometric air-fuel ratio.

Here, the value of the air-fuel ratio feedback correction coefficient α is changed by proportional-plus-integral (PI) control for stable control.

That is, comparing the O ₂ sensor output with the slice level,
When the air-fuel ratio is higher or lower than the slice level, the air-fuel ratio is not suddenly thickened or thinned, and when the air-fuel ratio is thick (thin), the air-fuel ratio is first lowered by P (raised), and then I The air-fuel ratio is controlled to be lighter (thicker) by gradually increasing it by the minute.

However, in a region where λ control is not performed, α = 1 is clamped and various correction factors COEF are set to obtain a desired air-fuel ratio.

By the way, if the base air-fuel ratio when α = 1 in the λ control region can be set to the stoichiometric air-fuel ratio (λ = 1), feedback control is not necessary, but in practice components (such as an air flow meter and a fuel injection valve) are required. , Pressure regulator, control unit), variations over time, non-linearity of fuel injection valve pulse width-flow rate characteristics, changes in operating conditions and environment, and other factors cause the base air-fuel ratio to deviate from λ = 1. Therefore, feedback control is performed.

However, if the base air-fuel ratio deviates from λ = 1, it takes time to set the step difference in the base air-fuel ratio to λ = 1 by feedback control when the operating region changes greatly. And for this purpose proportional and integral constants (P / I)
Since the amount is increased, overshoot or undershoot occurs and controllability deteriorates. That is, when the base air-fuel ratio deviates from λ = 1, the air-fuel ratio control is performed within a range that is considerably different from the theoretical air-fuel ratio.

As a result, the operation will be performed in a place where the conversion efficiency of the three-way catalyst is poor, and the cost will increase due to an increase in the amount of precious metal in the catalyst, and the catalyst will have to be replaced due to the deterioration of the conversion efficiency due to deterioration of the catalyst. There was a problem.

Therefore, by setting the base air-fuel ratio to λ = 1 by learning, it is possible to eliminate the deviation from λ = 1 caused by a step in the base air-fuel ratio at the time of transition, and to reduce the P / I component, making controllability easier. A learning control device for the base air-fuel ratio was devised to improve the fuel consumption and reduce the cost of the catalyst.

That is, a map of the learning correction coefficient αo corresponding to the engine operating conditions such as the engine speed and the load is provided on the RAM, and the basic injection amount Tp is corrected by αo as shown in the following equation when the injection amount Ti is calculated.

Ti = Tp × COEF × α × αo + Ts Then, the learning of αo proceeds in the following procedure.

i) In the steady state, the engine operating condition at that time and the control center value αc of α are detected.

ii) Search for αo that has been learned so far in correspondence with the engine operating conditions.

iii) The value of αo + Δα / M is obtained from αc and αo, and the result (learning value) is updated as a new αo and the memory is updated.

In addition, Δα indicates a deviation amount from the reference value α ₁ , and Δα = α
c-α ₁ , and the reference value α ₁ is generally 1.0. M is a constant.

By the way, in such a learning method in the conventional air-fuel ratio feedback control, since the deviation amount Δα cannot be detected accurately unless it is in the steady state, the learning is performed by detecting Δα only in the steady state. Then, in the transient operation state, learning is not performed in the operation area in which the operation is performed only temporarily.

Therefore, for example, as shown in FIG. 1, a hatched area in which the degree of progress of learning is large (hereinafter referred to as a learning area)
And other regions (hereinafter referred to as unlearned regions) in which the degree of progress of learning is small. Then, if the operating state changes in this state as indicated by the arrow, when an air-fuel ratio shift occurs in the system, the difference between the learning region and the unlearned region is α
Since there is a gap in the correspondence between the air-fuel ratio λ and the air-fuel ratio λ, there is a step in the air-fuel ratio when transitioning between the learned region and the unlearned region, leading to deterioration of exhaust emission in the transient state,
There was a problem that the effect of trying to learn was not practical.

<Object of the Invention> The present invention has been made in view of such conventional problems, and by using the estimated learning value by estimating the learning value in the unlearned area from the learning value in the learning area. An object of the present invention is to provide a feedback control device with a learning function capable of improving control accuracy in a transient operation state.

<Structure of the Invention> Therefore, according to the present invention, as shown in FIG. 2, according to the operating condition for each different operating region, the control value of the controlled object operated under the changing condition follows the target value. Control amount setting means for setting the control amount by adding or subtracting the feedback correction amount to the fixed control amount set by the above, and learning for correcting and updating the fixed control amount to reduce the deviation amount from the control center value of the feedback correction amount. In the feedback control device with a learning function, the learning control means includes a learning means for performing the learning, and storing the learning value. An auxiliary learning system that sets a learning value in a driving region with a small learning progress based on the determination from the determination means by interpolation calculation from a plurality of learning values in a driving region around the driving region and with a large learning progress. It will be configured with learning means.

<Examples> Examples will be described below.

The hardware configuration is shown in FIG.

1 is a CPU, 2 is a P-ROM, 3 is a CMOS-RAM for learning control,
Reference numeral 4 is an address decoder. For the RAM3, a backup power supply circuit is used to retain the stored contents even after the key switch is turned off.

As an analog input signal to the CPU 1 for controlling the fuel injection amount, an intake air flow rate signal from the hot wire air flow meter 5, a throttle opening signal from the throttle sensor 6,
There are a water temperature signal from the water temperature sensor 7, an exhaust oxygen concentration signal from the O ₂ sensor 8, and a battery voltage from the battery 9,
These are analog input interface 10 and A / D converter 1
It is supposed to be input via 1. Reference numeral 12 is an A / D conversion timing controller.

Digital input signals include ON / OFF from the idle switch 13, start switch 14 and neutral switch 15.
There are OFF signals, these are digital input interfaces 1
It is designed to be entered via 6.

In addition, the reference signal for every 180 ° and the position signal for every 1 ° from the crank angle sensor 17 are input through the one-shot multi-circuit 18. In addition, the vehicle speed signal from the vehicle speed sensor 19 is a waveform shaping circuit.
It is supposed to be input via 20.

Output signal from CPU1 (drive pulse signal to fuel injection valve)
Are sent to the fuel injection valve 22 via the current waveform control circuit 21.

Here, the CPU 1 controls the fuel injection amount by performing input / output operations and arithmetic processing according to a program (stored in the ROM 2) based on the flowchart (fuel injection amount calculation routine) shown in FIG.

Next, the flowchart of FIG. 4 will be described.

At S1, the intake air flow rate Q obtained by the signal from the air flow meter 5 and the engine speed N obtained by the signal from the crank angle sensor 17 are used to determine the basic injection amount Tp (= K × Q /
N) is calculated.

Set various correction factors COEF in S2.

In S3, the count value C of the update counter (which counts up in S15 described later) that counts the number of times the learning correction coefficient αo is updated is compared with a predetermined value. After decreasing / I by a predetermined amount, the process proceeds to S5. If it is less than the predetermined value, the process directly proceeds to S5 without changing the P / I amount.

In S5, the output from the O ₂ sensor 8 is compared with the slice level, and the air-fuel ratio feedback correction coefficient α is set by the proportional-plus-integral control based on the P / I component.

In S6, the voltage correction amount Ts is set based on the battery voltage from the battery 9.

In S7, the learning correction coefficient αo in the corresponding operating region is retrieved from the map described later based on the current engine speed N and the basic injection amount (load) Tp. The map of the learning auxiliary system coefficient αo for the engine speed N and the basic injection amount Tp is rewritable RAM3.
Are stored in the memory, and αo = 1 at all when learning is not started. Further, this map has N = 8 lattices and Tp = 4 lattices.

S8 to S11 are provided to detect the steady state, and S8
In 8 the change in vehicle speed is judged based on the signal from the vehicle speed sensor 19, in S9 the gear position is judged based on the signal from the neutral switch 15, and in S10 the throttle opening degree is judged based on the signal from the throttle sensor 6. The change is determined, and it is determined in S11 whether or not the predetermined time has elapsed. If it is within the predetermined time, the process returns to S8. Thus, if the change in vehicle speed is less than or equal to the predetermined value within the predetermined time, and the gear is engaged, and the change in the throttle opening is less than or equal to the predetermined value, it is determined that the steady state is reached, and in S12 to S14. The learning correction coefficient αo is corrected. Also, if the change in vehicle speed exceeds a predetermined value at any time within a predetermined time, if it becomes neutral, or if the change in throttle opening exceeds a predetermined value, it is determined to be in a transient state, The learning correction coefficient αo is not corrected in S12 to S14.

The correction of the learning correction coefficient αo in S12 when it is determined to be in the steady state is performed based on the mathematical expression αo ← αo + Δα / M as in the conventional case described above.

In S13, the new learning correction coefficient αo is written in the corresponding operation area of the map of RAM3. That is, the data in RAM3 is updated.

In S14, the count value C of the update number counter that counts the number of times the learning correction coefficient αo is updated in the current operation region is incremented.

In S15, the count value C of the update number counter in the current operation area is compared with a predetermined value C _1, and if C ≧ C ₁ and the learning progress is large, the process immediately proceeds to S18 and the injection amount is described later. Ti is calculated.

Among the operating region around the unlearned region C <a C ₁ learning progress in the judgment of S15 is in S16 if it is determined to be smaller, searches the learning area marked C ≧ C ₁ .

For example, in FIG. 1, when the driving region changes in the direction of the arrow, when it is in the unlearned region a, upper and lower learning regions A and B are searched on the map (if it is in the unlearned region b, the learning region is learned. Areas A, B, D are searched).

Next, at S17, the learning correction coefficient αo in the learning area, for example, A and B, retrieved as described above is read out, and the learning correction coefficient αo in the unlearned area a is calculated from these αo by proportional interpolation.

Next, in S18, the injection amount Ti is calculated by the following equation.

Ti = Tp × COEF × α × αo + Ts The injection amount Ti is calculated by the above, and a drive pulse signal corresponding to this injection amount Ti is transmitted via the current waveform control circuit 21 to the fuel injection valve.
It is given to 22 at a predetermined timing.

As for the learning area, the injection amount control can be performed with high accuracy by the learning correction coefficient αo actually learned in the same area as in the conventional case, and in the unlearned area, based on the learning correction coefficient of the surrounding learning areas. Since the injection amount control is performed using the highly reliable estimated learning correction coefficient obtained by the interpolation calculation, the difference in air-fuel ratio between the learned region and the unlearned region is eliminated, and the exhaust mission The deterioration can be prevented and the transient characteristics can be made smooth.

It should be noted that the estimated learning correction coefficient is stored in the map of the RAM, and based on the estimated learning correction coefficient and the latest value of α stored in the map when the learning area changes to C ≧ C ₁ and changes to the learning area. Alternatively, the learning correction coefficient αo may be updated.

<Effects of the Invention> As described above, according to the present invention, an estimated learning value is obtained by interpolation from a learning value in a region with a small learning progress and a region with a large learning progress, and feedback is performed based on the estimated learning value. Since it is configured to perform control, stable control characteristics can be obtained even in the transient operating state of the controlled object.For example, when applied to an air-fuel ratio feedback control device of an internal combustion engine, it is possible to improve the exhaust mission by eliminating the step of the air-fuel ratio. Features can be obtained.

[Brief description of drawings]

FIG. 1 is a graph for discriminating between a learning area and an unlearned area, FIG. 2 is a block diagram showing a configuration of the present invention, FIG. 3 is a hardware configuration diagram showing an embodiment of the present invention, and FIG. It is a flowchart which shows the control process of an Example same as the above. 1 ...... CPU, 3 ...... learning control for CMOS-RAM, 5 ...... air flow meter, 8 ...... O ₂ sensor, 17 ...... crank angle sensor, 22 ...... fuel injection valve

Claims (1)

[Claims] 1. A feedback control amount is added to or subtracted from a fixed control amount that is set according to operating conditions for different operating regions so that the control value of a controlled object operated under changing conditions follows the target value. Control amount setting means for setting a control amount by means of learning, learning for correcting and updating the fixed control amount to reduce the deviation amount between the control center value of the feedback correction amount and the reference value, and learning for storing the learned value In the feedback control device with a learning function, the learning progress determining means for determining the degree of progress of learning by the learning means is provided, and the learning progress is small based on the determination from the determining means. Learning by providing auxiliary learning means for setting the learning value of the driving region of the above-mentioned learning value by interpolation calculation from a plurality of learning values in the driving region around the driving region and having a high degree of learning progress Noh with feedback control device.