2022 Fiscal Year Final Research Report
Design of control systems using quaternion neural networks for dynamic systems
| Project/Area Number |
20K11980
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 61040:Soft computing-related
|
|---|
| Research Institution | Doshisha University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | 高次元ニューラルネットワーク / 四元数 / 適応・学習制御 |
|---|
| Outline of Final Research Achievements |
This study aims to establish a method for designing an adaptive/learning control of dynamical systems using a high-dimensional neural network based on quaternion algebra and to clarify the characteristics of using quaternion neural networks for control systems applications. Feedforward and recurrent quaternion neural networks and their learning algorithms are derived. Subsequently, a design method for servo-level controllers using quaternion neural networks is proposed and the stability analysis of the control system is also investigated. Furthermore, the feasibility and effectiveness of the proposed method are demonstrated via the computational experiments of controlling robot manipulators and non-linear systems for an application to real-world problems.
|
| Free Research Field |
制御工学
|
| Academic Significance and Societal Importance of the Research Achievements |
高次元ニューラルネットワークの一つである四元数ニューラルネットワークの特徴と能力を明らかにし，システムの制御への可能性を解明した成果は，人工ニューラルネットワークの研究において深層学習と並び注目されている高次元化に関する新たな知見として学術的意義があるとともに，計算知能の応用領域の拡大や制御工学分野における適応・学習制御手法の拡充に貢献することから，工業的有用性の点からも意義を有している．
|
