Control device, control method, and control program

Patent metadata

Field	Value
Publication number	US-11762344-B2
Application number	US-202017270841-A
Country	US
Kind code	B2
Filing date	Feb 25, 2020
Priority date	Feb 25, 2019
Publication date	Sep 19, 2023
Grant date	Sep 19, 2023

Abstract

The present invention reduces the probability of malfunction occurrence when performing predictive control of a device being controlled. In this control device of one aspect of the present invention, a prediction model for a control variable is used to calculate a prediction value from a measured value of the control variable, and a desired command value of the control variable is determined by correcting a desired basic value in accordance with the calculated prediction value. The degree of correction is determined on the basis of weight. The control device controls the operation of the device being controlled according to the determined desired command value. The control device assesses whether the device being controlled is operated appropriately on the basis of monitoring data relating to the operation result of the device being controlled, and optimizes the weight of the correction to make appropriate control possible based on the assessment result.

First claim

What is claimed is:

1. A control device comprising:
a first acquisition unit which acquires a desired basic value of a control variable;
a second acquisition unit which acquires a measured value of a control variable obtained by a first sensor measuring a control variable of a device being controlled;
a prediction unit which calculates a prediction value of the control variable from the acquired measured value using a prediction model of the control variable;
a desired correction unit which determines a desired command value of the control variable by correcting the desired basic value in response to the calculated prediction value and determines a degree of correction on the basis of a weight;
an operation control unit which controls an operation of the device being controlled in accordance with the determined desired command value;
a third acquisition unit which acquires monitoring data obtained by a second sensor monitoring an operation result of the device being controlled;
a result determination unit which assesses whether the device being controlled is appropriately operated by the control on the basis of the acquired monitoring data;
and a weight optimization unit which optimizes the weight of correction so that the device being controlled is appropriately operated on the basis of the assessment result;
wherein a plurality of candidates for the weight are given,
and wherein the weight optimization unit optimizes the weight of correction so that the device being controlled is appropriately operated by selecting a candidate of the weight having a highest probability that the device being controlled is appropriately operated as a result of controlling the operation of the device being controlled in accordance with each desired command value determined by a correction using the candidate of each weight.

2. The control device according to claim 1 ,
wherein the weight optimization unit optimizes the weight of correction to increase the probability that the device being controlled is appropriately operated by performing reinforcement learning based on a reward determined in response to the assessment result.

3. The control device according to claim 1 ,
wherein the desired correction unit calculates a difference between the desired basic value and the prediction value, determines a correction value on the basis of the calculated difference and the weight, and calculates the desired command value by adding the determined correction value to the desired basic value.

4. The control device according to claim 3 ,
wherein the weight comprises a proportional constant and a constant term, and
wherein the desired correction unit calculates a product of the difference and the proportional constant and adds the constant term to the calculated product to calculate the correction value.

5. The control device according to claim 1 ,
wherein the operation of the device being controlled by the control is repeated every takt cycle,
wherein the takt cycle comprises a first time interval and a second time interval after the first time interval,
wherein the first acquisition unit acquires a desired basic value of the control variable in the second time interval,
wherein the second acquisition unit acquires a measured value of the control variable in the first time interval,
wherein the prediction unit calculates the prediction value of the control variable in the second time interval from the measured value of the control variable in the first time interval using the prediction model,
wherein the desired correction unit determines the desired command value of the control variable in the second time interval by correcting the desired basic value in the second time interval in response to the calculated prediction value, and
wherein the operation control unit controls the operation of the device being controlled in the second time interval in accordance with the determined desired command value.

6. The control device according to claim 5 ,
wherein the prediction unit converts the measured value in the first time interval into a feature amount, inputs the converted feature amount to the prediction model, and performs a calculation process of the prediction model to acquire an output value output from the prediction model as the prediction value in the second time interval.

7. The control device according to claim 1 , further comprising:
a data collection unit which collects the measured value of the control variable obtained by the first sensor during control assessing that the device being controlled is appropriately operated as actual value data; and
a model update unit which updates the prediction model on the basis of the collected actual value data.

8. The control device according to claim 7 , further comprising:
a desired derivation unit which derives the desired value of the control variable for appropriately controlling the device being controlled on the basis of the collected actual value data,
wherein the first acquisition unit acquires the derived desired value as the desired basic value.

9. The control device according to claim 1 ,
wherein the operation control unit calculates a difference between the desired command value and the measured value, determines a value of the operation variable in response to the calculated difference, and gives the determined value of the operation variable to the device being controlled to control the operation of the device being controlled.

10. The control device according to claim 1 ,
wherein the device being controlled is a press machine comprising a mold,
wherein the control variable is a position of the mold,
wherein the first sensor is a position sensor configured to measure the position of the mold, and
wherein the second sensor is a camera disposed to photograph a workpiece pressed by the press machine.

11. A control method comprising:
acquiring a desired basic value of a control variable;
acquiring a measured value of a control variable obtained by a first sensor measuring a control variable of a device being controlled;
calculating a prediction value of the control variable from the acquired measured value using a prediction model of the control variable;
determining a desired command value of the control variable by correcting the desired basic value in response to the calculated prediction value and determining a degree of correction on the basis of a weight;
controlling an operation of the device being controlled in accordance with the determined desired command value;
acquiring monitoring data obtained by a second sensor monitoring an operation result of the device being controlled;
assessing whether the device being controlled is appropriately operated by the control on the basis of the acquired monitoring data; and
optimizing the weight of correction so that the device being controlled is appropriately operated on the basis of the assessment result;
wherein a plurality of candidates for the weight are given,
and wherein the optimizing the weight of correction so that the device being controlled is appropriately operated by selecting a candidate of the weight having a highest probability that the device being controlled is appropriately operated as a result of controlling the operation of the device being controlled in accordance with each desired command value determined by a correction using the candidate of each weight.

12. A non-transitory computer readable recording medium storing a control program allowing a computer to perform:
acquiring a desired basic value of a control variable;
acquiring a measured value of a control variable obtained by a first sensor measuring a contro

Assignees

• Omron Tateisi Electronics Co

Inventors

• Toku Takahiro
• Ueyama Yuki
• Inamoto Shuji
• Abe Yasuaki

Classifications

• G05B13/048 Primary

  using a predictor · CPC title

• G05B13/041 Primary

  in which a variable is automatically adjusted to optimise the performance · CPC title

• G05B13/044

  not using a perturbation signal · CPC title

• G05B13/0265

  the criterion being a learning criterion · CPC title

• B30B15/26

  Program-control arrangements · CPC title