JP3210136B2 - Equipment control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device control device.

[0002]

2. Description of the Related Art Recently, in order to evaluate the control of a device without actually moving the device, the device is modeled, and this modeled device model is composed of the same control logic as the control logic for the device. A combination with a control model has been proposed (Japanese Unexamined Patent Publication No.
59439).

[0003]

By the way, recently,
In addition to the equipment and the actual control system that actually controls the equipment, a simulation system using the above-described equipment model is provided, and the simulation results are reflected in the actual control system. Is thought to be performed better.

However, actual devices do not always exhibit the same dynamic characteristics due to individual differences and aging. Therefore, even if the result obtained by the simulation is reflected in the actual control system, the actual control of the device may not always be properly performed due to the difference between the actual device and the device model. It will be easy.

SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a device control apparatus in which a device model always matches the dynamic characteristics of an actual device so that the device can be controlled better. .

[0006]

Means for Solving the Problems To achieve the above object, the present invention has the following configuration. That is, in a device control device including a device and an actual control system that actually controls the device, a first device in which the dynamic characteristics of the device are modeled based on the correspondence between the input and output of the device. And controlling the first device model with the same control logic as the control logic of the real control system during the control of the device by the real control system and the model and the second device model. Identification means for adjusting the dynamic characteristic of the first device model so that the correspondence between the input and the output in the first device model matches the correspondence, and during the control of the device by the real control system, The second device model is controlled using the same control logic as the control logic of the actual control system, and the second model is controlled so that the output value of the device matches the output value of the second device model. Control ro Tsu by optimizing the control constants of the click, and norms means for changing the control constant of said actual control system to the optimized control constants, by the norm means the control constants of the actual control system
Before changing to the optimized control constant,
Controlling the first device model using an optimized control constant
To determine whether the control constant is appropriate.
Determining means, and the optimized
If it is determined that the control constant is not appropriate,
Change the control constants of the system to the optimized control constants
And a prohibition means for prohibiting the operation .

Preferred embodiments of the present invention based on the above configuration are as described in claims 2 and 3 in the claims.

[0008]

According to the first aspect of the present invention,
Differences or changes in the dynamic characteristics of the equipment due to individual differences or aging are compensated for by adjusting the dynamic characteristics of the first equipment model by the identification means, and the simulation results obtained using the first equipment model are actually used. As in the case of the simulation using the above device, the control is always appropriate and the control of the device by the actual control system in which the result of the simulation is reflected can be made more appropriate. In addition, the actual control system that actually controls the device can be managed within the performance range guaranteed by the second device model, while changing the control constants by the reference means to make it appropriate.
Further, before the optimized control constant is reflected in the actual control system, it is verified in advance using the first device model whether or not the optimized control constant is appropriate. Can be prevented from becoming inappropriate.

According to the second aspect of the present invention, when the control constant is not properly adjusted by the reference means, the change of the control constant by the reference means is prohibited. This prevents a problem in identifying the first device model.

[0010]

[0013] The configuration as described in claim 3 is preferable in that the control for setting the output value of the device to a predetermined target value is performed well.

[0012] With the configuration as described in claim 4 , it is possible to favorably perform the control for setting the idle speed of the engine to the target idle speed.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to the accompanying drawings. In the embodiment, idle speed control (ISC) is used, and the device to be controlled is an ISC valve for adjusting the intake air amount.

In FIG. 1, reference numeral 1 denotes an intake passage of an engine, in which an air cleaner 2, an air flow meter 3, and a throttle valve 4 are sequentially arranged from the upstream side to the downstream side.
A bypass passage 6 that bypasses the throttle valve 4 is provided in the intake passage 1, and an ISC valve 7 is connected to the bypass passage 6 as idle speed adjustment means.

The ISC valve 7 adjusts the amount of intake air passing through the bypass passage 6 to adjust the idle speed. The opening of the ISC valve 7 is continuously variable by the electromagnetic actuator 5. It has become so. That is, during the idling operation in which the accelerator opening is zero and the engine speed is equal to or lower than the predetermined speed, the ISC valve 7 is controlled such that the engine speed becomes the target idle speed (for example, 700 rpm). This ISC valve 7
That is, a control unit for controlling the actuator 5 is indicated by U, and sensors or switches for picking up signals used for control by the control unit U are indicated by S. The output value from the control unit U to the actuator 5 is set as a duty ratio as a control value for setting the target idle speed.

FIG. 2 is a block diagram showing the control contents of the control unit U, and includes each part except for a part 21A shown as a real engine in the figure. In FIG. 2, what is indicated roughly by reference numeral U1 is an actual control system that controls the actual engine 21A. Further, a model indicated by reference numeral UX is a model system for identification, and includes a first device model 21B and a first device model for controlling the same.
And a model control system U2. Further, the code UY
Is a reference model system, which is composed of a second device model 21C and a second model control system U3.

The actual control system U1 includes an integration circuit 22A for feedback control and an observer circuit 23A which is a main component of modern control. The target idle speed NT is input to the actual control system U1, and the
By A, the deviation between the NT and the actual engine speed (idle speed) NE1 is input to the integration circuit 22A.

The observer circuit 23A calculates a predetermined control value based on the actual engine speed NE1 and the input value to the actual engine 21A, that is, the duty ratio to the actuator 5. Then, the difference between the control value calculated by the observer circuit 23A and the output value from the integration circuit 22A is calculated by the subtractor 25A , and the calculation result is used as the input value to the actual engine 21A (observer). Input value to the buffer circuit 23A).

The first model control system U2 is a first device model (hard model) which models the dynamic characteristics of an actual engine.
21B is controlled. The first device model 21B is set based on the correspondence between the input value of the real engine 21A and the output value, and in a state where the dynamic characteristic of the real engine 21A completely matches, They are set to have the same output value. The first model control system U2 for the first device model 21B includes an actual control system U1
The control logic is set to be the same as that described above, and the components corresponding to the components in the actual control system U1 are indicated by using the symbol “B” instead of the symbol “A” in the actual control system U1. is there. The target idle speed as an input value in the real control system U1 and the first model control system U2 is shared as NT, and the output value in the real control system U1 is the engine speed NE1 in the real engine 21A.
The output value of the model control system U2 is the calculated engine speed NE2.

The second model control system U3 has the same configuration as the first model control system U2, and the second device model 21C has the same configuration as the first device model 21B. However, while the first device model 21B is identified corresponding to the real engine 21A and its dynamic characteristics are changed,
The device model 21C is for reference, and its dynamic characteristics are unchanged. The target idle speed NT is also input to the second model control system U3, while the calculated output value of the second device model 21C is set to NE3.

The control system U1 and the model systems UX and UY are:
It is managed by the management circuit 26, and the first
Identification control for adjusting the dynamic characteristics of the device model 21B and optimization of the control constants of the actual control system U1 are performed. For this reason,
Identification circuit 2 for adjusting dynamic characteristics of first device model 21B
8. An adjusting circuit 27 for adjusting the control constant of the actual control system U1 is provided, and both circuits 27 and 28 are operated under the control of the management circuit 26.

FIG. 3 shows the first device model 21 shown in FIG.
13B shows a specific example of B (the same applies to the second device model 21C). The circuit R3 is for obtaining the magnitude of the engine generated torque, and the charging amount Q, the fuel injection amount TP, and the ignition timing IG are used as parameters for this purpose. In order to obtain the filling amount Q used in the circuit R3, each processing in the circuits R1, R2, and R4 is performed.
TVO in the circuit R1 is the throttle opening, D in the circuit R2
"uty" indicates the duty ratio to the actuator 5. The outputs from the circuits R1 and R2 are added by an adder R8, subjected to a delay process in a circuit R4, and then filled with a charge amount Q.
Is output to the circuit R3.

The circuit R6 indicates the torque loss of the engine, and includes charging efficiency and pumping loss data. The loss torque TH in the circuit R6 is subtracted by the subtractor R9 from the torque calculated in the circuit R3, and the torque T after the subtraction is input to the circuit R7. The circuit R7 calculates the engine speed NE2 in the model control system U2 according to the equation shown here. In addition,
In the equation indicated by R7, I and K are control constants.

The circuit 22B in the first model control system U2 (same for the second model control system U3) shown in FIG.
Each characteristic formula of 23B is shown in FIG. In FIG. 4, "i" is a suffix, and KI is an integrating circuit 22.
B is a control constant (integral constant), and K1 to K7 are control constants of the observer circuit 23B. In the characteristic equation of the observer circuit 23B, the present value and the previous value of the output value NE (NE2) and the duty ratio of the output value NE to the actuator 5 up to five times before the previous value are used. Seven values are used as calculation parameters.

Here, the circuit 22 in the actual control system U1
The characteristic expressions at A and 23A are shown in FIG.
It is set in the same manner as the characteristic formula of 23B. In other words, the characteristic expressions of the circuits 22B and 23B (22C and 23C) are set according to the circuits 22A and 23A.

The control unit U is configured to include a real control system U1, an identification model system UX, a reference model system UY, and circuits 26 and 27 in FIG. 2 using a microcomputer. About the control contents,
This will be described with reference to the flowcharts shown in FIGS. The control described below is performed on the assumption that the vehicle is stopped.

First, at Z1 in FIG. 5, it is determined whether or not the reference control is not performed well and the reference is not good. This determination will be clear from the description of the reference control described later. Even if the control constant of the second model control system U3 is repeatedly optimized a predetermined number of times, the output value NE3 of the second device model U3 is obtained.
This is when the difference between the output value NE1 and the output value NE1 of the device 21A does not decrease. When the determination of Z1 is NO, whether or not the identification model system is converging in Z2,
It is determined whether or not the dynamic characteristics of the device model 21B completely or almost completely match the dynamic characteristics of the device 21A. If the determination of Z2 is NO, in Z3 to Z5, as described below, identification control for identifying the first device model 21B, high-speed prediction control using the identification model system UX, and real control system U1 Normative control for optimizing the control constants of

If the determination in Z1 is YES or the determination in Z2 is YES, the processing of Z6 and Z7 is performed. In this case, although the above-described identification and high-speed prediction control are performed, Normative control is not performed (prohibited).

FIG. 6 shows the contents of the control of the identification in the Z3 or Z6. The control constants of the first model control system U2 are, for example, adjusted by using the adjustment circuit 27 and the actual control system U1.
Is set to a state consistent with the control constant of. Z31 in FIG.
In, it is determined whether or not many inputs have changed for the actual engine 21A. Specifically, the amount of change in the throttle opening is equal to or more than a predetermined value, the amount of change in DFB indicating the amount of change in the duty ratio output to the actuator 5 is equal to or more than a predetermined value, and the target rotational speed NT When the three conditions that the amount of change of is equal to or more than a predetermined value are satisfied,
Since multiple inputs are being changed, the determination of Z31 is YES, and in this case, the
Is performed.

If the determination in Z31 is NO, it is determined in Z32 whether the error in the steady state, that is, the evaluation value Hi indicating the degree of coincidence between the actual engine 21A and the first equipment model 21B in the steady operation state is small. Is done. As described later, when the evaluation value Hi is small, the degree of coincidence is high. If the determination in Z32 is NO, in Z33, it is determined whether or not the actual engine 21A is currently in steady operation. If the determination in Z33 is YES, the processing in Z34 to Z37 identifies the first device model 21B during normal operation. The identification during the steady operation is performed by optimizing a control constant such as a time constant in the circuits R1, R2, and R3 shown in FIG .

At Z34, one combination is selected as the control constants for the circuits R1 to R3 from the first to nth stored combinations from the design of experiment map.
Next, in Z35, the selected i-th (i = 1 to 1)
The output value NE2 of the model control system U2 and the output value NE1 of the actual control system U1 obtained by the operation based on the combination of n)
The integral of the square of the absolute value of the deviation from
Evaluation value H indicating the degree of error for the i-th integration constant
i is determined. As described above, the smaller the evaluation value Hi, the more preferable.

The evaluation value Hi at Z35 is sequentially obtained for the first to nth values, and the result is stored as H1 to Hn at Z36. Thereafter, in Z37, the control constant that has indicated the smallest evaluation value among the evaluation values H1 to Hn stored in Z36 is the first constant.
It is used as a control constant for the circuits R1 to R3 of the device model 21B.

If the determination in Z32 is YES , Z3
At 8, it is determined whether or not the operating state of the actual engine 21A is in a transient state. Specifically, when the change amount of the engine speed NE1 is equal to or more than a predetermined value or when the change amount of DFB is equal to or more than a predetermined value, the determination of Z38 becomes YES.
In this case, the equipment model 21B at the time of transition is identified by the processing of Z39 to Z42. The identification during this transition optimizes the control constants of the circuits R4 and R5 in FIG . Note that this optimization method is performed substantially in the same manner as in the steady state,
The duplicate description is omitted.

If the determination in Z33 is NO , or if the determination in Z38 is NO, return is made without identification. It should be noted that it is conceivable to shift to Z38 when NO is determined in the determination of Z33.
To give priority to the identification of the control constants for R3 (to identify them first), if Z33 is NO, the
I have to make it.

FIG. 7 shows the contents of Z4 or Z7 in FIG. The contents in FIG. 7 are for predicting the possibility of engine stall. First, in Z51, it is determined whether or not the flag F has been changed to 0 or 1. The flag F becomes β2 (β1 is β2) after the change amount (absolute value) of the throttle opening reaches a predetermined value β1 (> 0).
2T (T)
Is set to 1 only for a time constant) seconds.

If the determination in Z51 is YES, in Z52, the time is set as a parameter and the throttle opening TV
An O prediction input pattern is created. Thereafter, in Z53, it is determined whether or not the high-speed operation using the identification model system UX to be executed in Z54, that is, the high-speed operation of the output value NE2 has been performed a predetermined number of times n or more. Note that n is an integer value obtained by rounding the calculated value of “2T / ST” (T is the above-described time constant, and S is a sampling period). If the determination in Z53 is NO, NE2 is calculated at a high speed in Z54 as described above, and the process returns to Z53 again.

If the determination in Z53 is YES, in Z55, the minimum value of NE2 calculated for 2T seconds is selected. Then, at Z56, the minimum N
It is determined whether E2 is smaller than 500 rpm.
If the determination in Z56 is YES, there is a risk of engine stall, and in this case, a process for increasing the intake air amount is performed in Z57 (the ISC valve 7 is corrected in the opening direction). Although the processing of Z57 is omitted in FIG. 2, for example, the subtractor 2 in FIG.
Immediately after 5A, a signal corresponding to the intake air amount increase correction may be input (added). If the determination in Z56 is NO, there is no danger of engine stall, and in this case, it is returned as it is.

If the determination in Z51 is NO, it is determined that it is not necessary to perform a high-speed engine stall prediction calculation.
At 8, the calculation of the identification model system UX is executed at a normal speed every ST seconds. The processing of Z58 is the same as that of Z
A standby state for performing the processing after 52 is entered.

FIG. 8 shows the contents of Z5 in FIG. In Z60 of FIG. 8 , a reference model system UY is used.
Optimization of control constants is performed (specific examples of optimization of control constants)
In the next Z61, it is determined whether the identification error is small, that is, the output value NE1 of the actual machine and the identification model system U
It is determined whether or not the output value NE of X completely or almost completely matches. If the determination in Z61 is YES, the identification model system UX is used in Z62.
Of the control constants optimized using the reference model system UY
Verification is performed. Thereafter, the process proceeds to Z63, and the Z63
In the above, the control constants optimized as described later
It is determined whether the evaluation value He is small (He is small).
The evaluation is much higher). When the determination of Z63 is YES
Are the control constants KI, K of the actual control system U1 in Z64.
1 to K7 are changed to optimized control constants
When the determination of Z63 is NO, the process returns to
Changing the control constant of the control system U1 is prohibited. one
On the other hand, when the determination of Z61 is NO, the verification of Z62 is performed.
Is transferred to Z63 without any
NE1 is almost the same as NE2 or NE3
May be determined).

[0040]

[0041]

If the determination in Z61 is YES, the control constants are optimized using the reference model system UY.
Thereafter, verification of the optimized control constant using the identification model system UX may be performed, and the process may be shifted to Z64 only when there is no problem even if the optimized control constant is used. In this case, If the determination in Z61 is NO, the control may be directly returned to prohibit the optimization of the control constants (if the optimized control constants cannot be verified by the identification model system UY, the actual control system U1 may be disabled). Do not change the control constant).

FIG. 9 shows the contents of Z60 . In FIG. 9, prior to the control, the target idle speed is slightly increased stepwise for a predetermined period. First, Z
At 11, it is determined whether or not the optimization of the control constant KI for the integration circuit 22C has been completed, but initially this determination is NO and the process proceeds to Z12. In Z12, any one of the stored setting values from 1 to n is selected for the integration constant KI from the experiment design map. Then, the selected i-th (i = 1 to n)
Of the absolute value of the deviation between the output value NE2 of the first model control system U2 and the output value NE1 of the actual control system U1 obtained by the operation based on the set value of Evaluation value Hei indicating the degree of error for a constant
Is determined. The smaller the evaluation value Hei , the better .

The evaluation value He at Z13 is sequentially obtained for the integration constants from the first to the nth, and the result is Z
At 14, they are stored as He1 to Hen . After this,
In Z15, the evaluation value He1 stored in Z14
To Hen , and the evaluation value that indicates the minimum evaluation value and the corresponding integration constant KI are selected and set. Of course, the minimum evaluation value He selected and set in Z15 is used for the determination of Z63 in FIG. 8, and the integration constant KI is used in Z64 in FIG.
The integration constant KI selected and set in Z15 is used as the integration constant in the integration circuit 22A in the actual control system U1.

If the determination in Z11 is YES, in Z16, it is determined whether the evaluation value HE indicating the degree of error selected and set in Z15 is larger than a predetermined value. If the determination in Z16 is YES, the integration circuit 22
Since a sufficient evaluation cannot be obtained only by optimizing the integration constant KI in A, in this case, the processing after Z17 is performed to optimize the control constant of the observer circuit 23A. Since the processing of Z17 to Z20 is performed in the same manner as the processing of Z12 to Z15 only because the control constants are different, the duplicate description will be omitted. Of course, the control constants K1 to K7 selected and set in Z20
Are used as control constants for the observer circuit 23A in Z64 of FIG. If the determination in Z16 is NO, the process proceeds to Z12.

Although the embodiment has been described above, the present invention can be used not only for idle speed control but also for appropriate engine control such as air-fuel ratio control and supercharging pressure control. Appropriate control of the vehicle, such as traction control and ABS control for setting the slip value of the drive wheel to the target slip value, suspension control for setting the vehicle height to a predetermined target vehicle height, and 4WS control for setting the rear wheel to the target turning angle, etc. Applicable. Further, the present invention can be applied to control of appropriate devices other than the vehicle and control having no target value.

[Brief description of the drawings]

FIG. 1 is a diagram showing an idle speed adjustment portion to which the present invention is applied.

FIG. 2 is a block diagram showing a control system to which the present invention is applied.

FIG. 3 is a diagram showing an example of a device model corresponding to an actual engine.

FIG. 4 is a diagram showing a setting example of a control characteristic equation in a model control system.

FIG. 5 is a flowchart showing a control example of the present invention.

FIG. 6 is a flowchart showing a control example of the present invention.

FIG. 7 is a flowchart showing a control example of the present invention.

FIG. 8 is a flowchart showing a control example of the present invention.

FIG. 9 is a flowchart showing a control example of the present invention.

[Explanation of symbols]

 1: Intake passage 7: ISC valve (for adjusting idle speed) 5: Actuator 21A: Actual engine (equipment) 21B: First equipment model (for identification) 21C: Second equipment model (for reference) 26: Management Circuit 27: Adjustment circuit (for changing control constant) 28: Identification circuit U: Control unit U1: Actual control system UX: Model system for identification UY: Model system for reference U2: First model control system (for identification) U2: No. 2 model control system (for norm)

Claims (4)

(57) [Claims] 1. An apparatus control apparatus comprising an apparatus and an actual control system for actually controlling the apparatus, wherein a dynamic characteristic of the apparatus is modeled based on a correspondence between an input and an output of the apparatus. A first device model and a second device model; and controlling the first device model with the same control logic as the control logic of the real control system during control of the device by the real control system. Identification means for adjusting a dynamic characteristic of the first device model such that a correspondence relationship between an input and an output in the first device model matches a correspondence relationship with an output; and control of the device by the real control system. The second device model is controlled by the same control logic as the control logic of the actual control system so that the output value of the device matches the output value of the second device model. The system that controls Control means for optimizing the control constants of the control logic and changing the control constants of the actual control system to the optimized control constants; and controlling the control constants of the real control system by the reference means.
Before changing to the specified control constant,
Controlling the first device model using the controlled constants
To determine whether the control constant is appropriate or not
The optimized control constant is appropriate by the step and the determination means.
If not, the control constant of the actual control system is changed.
Prohibition of changing to the optimized control constant
Means for controlling a device. 2. The method according to claim 1, wherein even if the control value is optimized by the reference device, if the output value of the device and the output value of the second device model are significantly different from each other, the reference device uses the control device. A device further comprising a prohibition means for prohibiting a change of a control constant. 3. The apparatus according to claim 1, wherein the control of the device by the actual control system is performed such that an output value of the device becomes a predetermined target value. 4. The apparatus according to claim 3, wherein the device is idle speed adjusting means for adjusting an idle speed of the engine, and the target value is a target idle speed.