JP-2026075908-A - Calibration device, calibration method, and calibration program

Abstract

[Problem] To easily maintain the accuracy of automatic control of the device. [Solution] In the calibration device 10, the calculation unit 15b calculates a calibration value for the control value based on the deviation between the measured value and the target value of the evaluation target in the device being processed, which is controlled by the control value. The addition unit 15c adds the calculated calibration value to the recommended value of the control value output by the model 14a, which has been learned by imitation learning. [Selection Diagram] Figure 8

Inventors

• 藤井 沙苗
• 伊藤 浩二

Assignees

• ＮＴＴドコモビジネス株式会社

Dates

Publication Date

20260511

Application Date

20241023

Claims (7)

1. A calculation unit calculates a calibration value for the control value based on the deviation between the measured value and the target value of the evaluation target in the device being processed, which is controlled by the control value. An adder adds the calculated calibration value to the recommended value of the control value output by the model learned through imitation learning, A calibration device characterized by having the following features.
2. The calibration device according to claim 1, characterized in that the calculation unit calculates the calibration value using the average value of the deviations over a predetermined past period.
3. The calibration apparatus according to claim 2, characterized in that the calculation unit calculates the calibration value by multiplying the average value of the deviations by a predetermined coefficient.
4. The calibration apparatus according to claim 1, characterized in that, if the absolute value of the calculated calibration value is greater than a predetermined threshold, the addition unit corrects the calibration value so that the increase or decrease range of the calibration value becomes the threshold, and then adds the calibration value to the recommended value of the control value.
5. The calibration device according to claim 1, characterized in that the addition unit adds the calibration value calculated at a predetermined period to the recommended value of the control value.
6. A calibration method performed by a calibration device, A calculation step of calculating a calibration value for the control value based on the deviation between the measured value and the target value of the evaluation target in the device being processed, which is controlled by the control value, An addition step is to add the calculated calibration value to the recommended value of the control value output by the model learned through imitation learning, A calibration method characterized by including the following:
7. A calculation step of calculating a calibration value for the control value based on the deviation between the measured value and the target value of the evaluation target in the device being processed, which is controlled by the control value, An addition step is to add the calculated calibration value to the recommended value of the control value output by the model learned through imitation learning, A calibration program to get a computer to run something.

Description

This invention relates to a calibration device, a calibration method, and a calibration program. Traditionally, PID (Proportional Integral Derivative) control has often been used for the automatic control of equipment in factories. PID control is a method of calculating the output value using three elements: the deviation between the measured value and the setpoint related to the controlled object, its integral, and its derivative (see Patent Document 1). In PID control, if the delay time (dead time) between the control operation and the change in the output value is long, PID control may not be applicable, and manual operation by a human may be necessary. In such cases, imitation learning, which involves learning past operation history (history of operating data indicating the situation, and history of control values indicating what changes were made to setpoints and output values), makes it possible to reproduce human operation and achieve automatic control of the device. When reproducing human actions, it becomes possible to mimic human actions regardless of environmental changes by using JIT (Just In Time) retraining based on past history of actions performed in situations similar to the one being reproduced (see Non-Patent Literature 1). Japanese Patent Publication No. 2019-185194 “Just-In-Time Predictive Control: Predictive Control Based on Accumulated Data,” [online], 2013, Measurement and Control, Vol. 52, No. 10, October 2013 issue, [Retrieved September 26, 2024], Internet <URL: https://www.jstage.jst.go.jp/article/sicejl/52/10/52_878/_pdf/-char/ja> Figure 1 is a diagram illustrating the overview of the calibration device.Figure 2 is a diagram illustrating the overview of the calibration device.Figure 3 is a diagram illustrating the overview of the calibration device.Figure 4 is a diagram illustrating the overview of the calibration device.Figure 5 is a diagram illustrating the overview of the calibration device.Figure 6 is a diagram illustrating the overview of the calibration device.Figure 7 is a diagram illustrating the overview of the calibration device.Figure 8 is a schematic diagram illustrating the general configuration of a calibration device.Figure 9 is a diagram illustrating the processing of the calculation unit.Figure 10 is a diagram illustrating the processing of the addition section.Figure 11 is a flowchart illustrating the calibration process.Figure 12 illustrates a computer running a calibration program. Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. However, the present invention is not limited to this embodiment. Furthermore, in the drawings, identical parts are denoted by the same reference numerals. [Overview of Calibration Device] Figures 1 to 7 are diagrams illustrating the overview of the calibration device. In a PID control device, the output value is determined by three elements: the deviation between the measured value and the set value of the controlled object, its integral, and its derivative. For example, as illustrated in Figure 1, PID control is performed for each controlled object, such as temperature and pressure. Operators may manually change the set values and output values of each PID control so that the evaluation target value related to the plant's operating target (e.g., impurity concentration) reaches a desired state (e.g., stabilizing the impurity concentration around 1%) (hereinafter, the set values and output values will be referred to as control values). Furthermore, automatic control that replicates human operation is achieved through imitation learning, which involves learning past operation history (history of driving data indicating the situation, and history of control value changes indicating what operations were performed). In other words, the controlled device is PID controlled using the control values calculated by the AI that has undergone imitation learning. Specifically, PID control of a controlled device using imitation learning has two modes, as illustrated in Figure 1: guidance mode and autopilot mode. Guidance mode, as illustrated in Figure 2(a), is a manual operation mode where the operator manually controls the device based on recommended control values calculated by the AI. In contrast, autopilot mode, as illustrated in Figure 2(b), is an automated operation (automatic control) mode where the device is directly controlled using recommended control values calculated by the AI. As illustrated in Figure 3(a), when the condition of the equipment changes due to seasonal variations, aging, changes in production volume, repairs, etc., a distribution shift occurs in the distribution of the explanatory variables (operational data) and the objective variable (control values), changing from the time of training. This leads to a decrease in the accuracy of the AI that has been trained through imitation. Therefore, as illustrated in Figure 3(b), the accuracy of the AI, which mimics hum