JPH06289906A - Adaptive control method and its device - Google Patents

Abstract

PURPOSE:To insert an inverse function and to execute highly precise control in spite of the presence or absence of the fluctuation of a dead zone by executing linear approximation for respective sections obtained by dividing the range of control input, obtaining respective functions and executing an adaptive rule processing with the function corresponding to the section to which control input is applied. CONSTITUTION:The range which control input can take is previously divided into the plural sections and the respective functions obtained by linearlly approximating the respective sections are held in a state where they are made to correspond to the respective sections. The section s are the section corresponding to the dead zone or a saturation area, and the section positioned in the both sections, for example. A system waits until output and a target value are received in a step (hereafter omitted) SP1, and control input is calculated in SP2. The section to which the control input belongs is detected in SP3. In SP4, the function corresponding to the detection section is selected so as to execute the adaptive rule processing. In SP5, the control rule processing is executed based on the result so as to repeat the processing of SP1 again. Thus, the dispersion of correction width for the respective sections can be suppressed even if the dead zone fluctuates, and highly precise control can be executed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an adaptive control method and an apparatus thereof, and more particularly to an adaptive control method and an apparatus thereof suitable for controlling a control object having a dead zone.

[0002]

2. Description of the Related Art Conventionally, adaptive control methods have been used in various fields in which processing based on an adaptive law is performed based on an output from a controlled object, and input control is performed based on a control law based on a result of the adaptive law processing. It is being adopted. Here, the process based on the adaptive law is, for example, the parameter identification process of the control target, and the process based on the control law is, for example, the process of calculating the gain of the controller based on the identified parameter.

Therefore, when the characteristic of the controlled object is non-linear and the parameter of the controlled object fluctuates, the adaptive control can be satisfactorily controlled by applying the adaptive control. Depending on the controlled object, there may be a dead zone (the area where the output does not change even if the control input changes), and this is input to compensate the dead zone in order to apply adaptive control to such a controlled object. There has been proposed a method of inserting an inverse function of a function defining an output relation into an output part of a controller. An example of such a control target is a hydraulic device.

FIG. 5 is a block diagram schematically showing the configuration of a control system in which the above-mentioned adaptive control method is implemented. The control target value is supplied to the controller 51, and the output of the controller 51 is the inverse function processing section 52. And the output value from the inverse function processing unit 52 is supplied to the controlled object 53. Then, the output value from the controlled object 53 is directly fed back to the controller 51. Further, the output value from the controller 51 and the output value from the controlled object 53 are both low-pass filter 5
It is supplied to the identifier 56 via 4, 55, and the gain of the controller 51 is controlled based on the identification output from the identifier 56.

[0005]

However, since the method of inserting the inverse function into the output part of the controller cannot adapt to the fluctuation of the function including the dead zone, effective adaptive control cannot be achieved. It will be described in more detail. In the above method, it is necessary to obtain the inverse function of the changed function in real time when the variation of the function, especially the variation of the dead zone occurs, but it is almost impossible to obtain the inverse function in real time. Therefore, as the inverse function, the adaptive control is performed using only the preset inverse function. As a result,
Dead zone compensation will be insufficient. Specifically, the dead zone fluctuation includes a dead zone shift and a dead zone width change.In the case of a hydraulic system, at least one of the fluctuations is caused by valve replacement, load fluctuation, temperature fluctuation, etc. Occurs. When any variation occurs, the inverse function of the function before the variation is simply inserted in the output part of the controller, and therefore the exact inverse function at the relevant time is different. Therefore, the control becomes inaccurate due to the deviation from the exact inverse function.

In other words, when the above-mentioned conventional method is adopted, when the dead zone fluctuates, a range in which the correction width is large and a range in which the correction width is relatively small coexist. Regardless of the variation of the correction width, the control accuracy is lowered due to the use of the preset inverse function as it is.

[0007]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and it is an adaptive control method and apparatus capable of achieving adaptive control for a control object having a dead zone with high accuracy regardless of whether the dead zone fluctuates. Is intended to provide.

[0008]

In order to achieve the above object, the adaptive control method according to claim 1 divides a control input range into a plurality of sections, and a control input and an output for each section. This is a method of setting a function that defines the relationship with, detecting the section to which the control input belongs, and performing adaptive control based on the function corresponding to the detected section.

According to another aspect of the adaptive control apparatus of the present invention, a dividing means for dividing the range of the control input into a plurality of sections, and a function holding means for holding a function for defining the relationship between the control input and the output for each section. And section detecting means for detecting a section to which the control input belongs, and function selecting means for selecting a function corresponding to the detected section as a function for adaptive control.

[0010]

According to the adaptive control method of claim 1, in performing adaptive control on a control object having a dead zone, the range of control input can be divided into a plurality of sections, and
Since a function that defines the relationship between control input and output is set for each section, the section to which the control input belongs is detected, and adaptive control based on the function corresponding to the detected section is performed, so the dead zone fluctuates. Even in such a case, the variation of the correction width for each section can be suppressed to a relatively small range, and high-precision control can be achieved.

According to another aspect of the adaptive control device of the present invention, the adaptive control is performed on the control target having a dead zone.
In advance, the range that the control input can take is divided into a plurality of sections by the dividing means, and the function holding means holds a function that defines the relationship between the control input and the output for each section. When actually performing adaptive control, the section detecting unit can detect the section to which the control input belongs, and the function selecting unit can select the function corresponding to the detected section as the function for adaptive control. . Therefore, even if the dead zone changes, the variation in the correction width for each section can be suppressed to a relatively small range, and high-precision control can be achieved.

[0012]

Embodiments will be described in detail below with reference to the accompanying drawings showing embodiments. 1 and 2 are flowcharts for explaining an embodiment of the adaptive control method according to the present invention. FIG. 1 illustrates preprocessing and FIG. 2 illustrates actual adaptive control processing.

In step SP1 of the flowchart of FIG. 1, the range that the control input can take is divided into a plurality of sections, each section is linearly approximated in step SP2, and each function obtained by the linear approximation is divided in step SP3. The section is held in a state of being associated with the section, and the series of processing is ended as it is. Therefore, for example, when the flow gain static characteristic of the hydraulic device to be controlled is as shown in FIG. 3, it is divided into a section corresponding to the dead zone, a section corresponding to the saturation region, and a section located between both sections. , By linearly approximating each section, the function indicated by the broken line in FIG. 3 can be obtained.

Wait until the output and the target value are received in step SP1 of the flowchart of FIG.
2, the control input is calculated, the section to which the control input belongs is detected in step SP3, the function corresponding to the detected section is selected in step SP4 to perform adaptive law processing, and control is performed based on the adaptive law processing result in step SP5. The rule process is performed, and the process of step SP1 is repeated again.

Therefore, even if the dead zone fluctuates as indicated by the alternate long and short dash line in FIG. 3, the difference between the function for each section and the actual characteristic does not significantly increase.
As a result, a decrease in parameter identification accuracy can be significantly suppressed, and high-accuracy adaptive control can be achieved.

[0016]

[Embodiment 2] FIG. 4 is a block diagram showing an embodiment of the adaptive control apparatus of the present invention, in which a controller 1 having a control target value and a feedback value as inputs, and an output value from the controller 1 as an input. Correction processing unit 2, a control target 3 to which the output value from the correction processing unit 2 is input, and a plurality of adaptive law processing units to which the output value from the controller 1 and the output value from the control target 3 are input 4 and a segmented section detection unit 5 which receives an output value from the controller 1 to detect which segmented section it belongs to and allows only the adaptive law processing unit 4 corresponding to the detected section to operate. is doing.

The correction processing section 2 holds the inverse function of each section in correspondence with the characteristic of the controlled object 3 including the dead zone divided into a plurality of sections. Then, the corresponding inverse function is selected based on the output from any of the adaptive law processing units 4 in which the operation is permitted. The adaptive law processing unit 4 performs parameter identification based on a function obtained by linearly approximating the corresponding section among the sections, and supplies a gain control signal corresponding to the parameter identification result to the controller 1. , And supplies an inverse function selection signal to the correction processing unit 2. The controller 1 inputs the control target value and the feedback value from the controlled object 3 and multiplies the difference between the two values by the gain corresponding to the parameter identification result.

The operation of the adaptive control device having the above configuration is as follows. To which section the output value from the controller 1 belongs belongs to the section detection unit 5, and only the adaptive law processing unit 4 corresponding to the detected section is allowed to operate. Therefore, in a state where the operation of any of the adaptive law processing units 4 is allowed, the same operation as that of the conventional adaptive control device can be performed.

However, in this case, since the adaptive processing corresponds to only one division section, even if the dead zone of the controlled object fluctuates, the range of the correction width largely fluctuates. Can be eliminated, and highly accurate adaptive control can be achieved. In this embodiment, the controller 1
Although the correction processing unit 2 and the correction processing unit 2 are configured separately, it is also possible to integrate the two, that is, to provide the controller with a correction processing function.

[0020]

As described above, according to the first aspect of the present invention, even when the dead zone changes, the variation of the correction width for each section can be suppressed to a relatively small range, and high precision control can be achieved. It has the unique effect of being able to do it. The invention according to claim 2 also has a unique effect that even if the dead zone changes, the variation in the correction width for each section can be suppressed to a relatively small range, and high-precision control can be achieved.

[Brief description of drawings]

FIG. 1 is a flowchart for explaining preprocessing in one embodiment of the adaptive control method of the present invention.

FIG. 2 is a flow chart for explaining actual adaptive control processing in one embodiment of the adaptive control method of the present invention.

FIG. 3 is a diagram showing an example of sections and linear approximation corresponding to a static flow rate gain characteristic of a hydraulic system.

FIG. 4 is a block diagram schematically showing an embodiment of the adaptive control device of the present invention.

FIG. 5 is a block diagram schematically showing a configuration of a conventional adaptive control device.

[Explanation of symbols]

 3 Control object 4 Adaptive law processing unit 5 Divided section detection unit

Claims (2)

[Claims] 1. An adaptive control method for a controlled object (3) having a dead zone, which divides a control input into a plurality of sections and sets a function for defining a relationship between the control input and the output for each section. An adaptive control method characterized by detecting a section to which a control input belongs and performing adaptive control based on a function corresponding to the detected section. 2. An adaptive control device for a controlled object (3) having a dead zone, said partitioning means (5) for partitioning a possible range of a control input into a plurality of sections, and a control input and an output for each section. Function holding means (4) for holding a function that defines the relationship between, the section detecting means (5) for detecting the section to which the control input belongs, and the function for selecting the function corresponding to the detected section as a function for adaptive control. An adaptive control device comprising: a selection means (5).