2023 Fiscal Year Final Research Report

Data driven control theory based on triad relationship on control, data and specification: Creation systematization and new evolution

Research Project

PDF

Project/Area Number	20H02169
Research Category	Grant-in-Aid for Scientific Research (B)
Allocation Type	Single-year Grants
Section	一般
Review Section	Basic Section 21040:Control and system engineering-related
Research Institution	The University of Electro-Communications
Principal Investigator	Kaneko Osamu 電気通信大学, 大学院情報理工学研究科, 教授 (00314394)
Co-Investigator(Kenkyū-buntansha)	定本知徳電気通信大学, 大学院情報理工学研究科, 准教授 (40839966) 増田士朗東京都立大学, システムデザイン研究科, 教授 (60219334) 山本豪志朗京都大学, 医学研究科, 准教授 (70571446)
Project Period (FY)	2020-04-01 – 2024-03-31
Keywords	データ駆動制御 / データ駆動予測 / 制御仕様 / 目標応答更新 / 定値制御 / 大規模系
Outline of Final Research Achievements	In the data-driven control, which is the design or update method of control by directly using the experimental data, we address the problem on control specification with expansion of data-driven prediction and data-driven control. Particularly, we give the update method of control design specification, which is deeply related to realizable performance of the plant and has been regarded as a given item for controller design, by using data-driven prediction and offline optimizations. Furthermore, we expand the data-driven control from the view point of the class of plant, control architecture, and situations. We also proposed a new data-driven control method based on data-driven prediction. Thus, we give many meaningful results from the both of practical and academic view points. In addition, we also give meaningful results on regulation problem and large scale systems from the viewpoint of data-driven control and learning.
Free Research Field	制御工学
Academic Significance and Societal Importance of the Research Achievements	対象の知識に基づいて与えられるとしていた制御仕様を，データから，オフライン最適化で実現可能なものを与える手法は，実応用で「どのような仕様を与えたらよいか」という頻繁に挙げられる問に対するデータ駆動制御からの解答を与えたことに相当し，実用的に意義がある．これにより，生産現場でも効率化がすすみ，大きなコストパフォーマンスが見込まれる．学術的にもデータ駆動予測の応用という点で重要な意義がある．また，従前まで拡張されていなかったクラスや状況への拡張などを行うことで，制御の一分野としてのデータ駆動制御の方法論をより理路整然と確立させた点にも，今後残されるであとろう学術的成果という点で意義がある．