Construction of a control system design method that utilizes the advantages of model-based design and data-driven design

2021 Fiscal Year Final Research Report

• Project/Area Number: 17K06496
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Control engineering/System engineering
• Research Institution: Hiroshima University
• Principal Investigator: Saeki Masami 広島大学, 先進理工系科学研究科(工), 名誉教授 (60144325)
• Co-Investigator(Kenkyū-buntansha): 和田 信敬 広島大学, 先進理工系科学研究科(工), 教授 (50335709)
• Project Period (FY): 2017-04-01 – 2022-03-31
• Keywords: データ駆動設計 / ロバスト制御 / 2自由度系 / ループ整形 / パラメータ空間設計 / H∞制御 / システムゲイン

Outline of Final Research Achievements

Data-driven design, in which a simple controller such as a PID controller is adjusted off-line using the plant's input-output response, is not suitable for unstable and oscillatory plants. To deal with this problem, a two-step method was proposed in which the plant is stabilized by model-based design and data-driven design is applied to the closed-loop system. In the model-based design, an implementation method of stabilized controllers that are suitable for data-driven design was presented. In data-driven design, a method for drawing PID gain sets in parameter space and a method for evaluating system gains using input-output response data were presented, and the conventional design method was extended to multi-input-output systems. Numerical experiments showed the usefulness of the two-step method.

Free Research Field

ロバスト制御系の解析と設計法およびデータ駆動設計法の研究

Academic Significance and Societal Importance of the Research Achievements

モデルベーズド設計の成果によりオーバーシュートの小さい目標値応答を与えるロバスト安定化制御器の簡便な実装が可能となり，データ駆動設計の成果によりステップ応答を用いたシステムゲインの下界の評価や既存のデータ駆動設計のゲイン調整法の多入出力系への拡張が可能となった。モデルベーズド設計を併用することでデータ駆動設計を不安定，非最小位相系などを含む幅広い制御対象に適用可能となった。