Covariance matrix adaptation evolution strategy based design of fixed structure robust H∞ loop shaping controller |
| |
| Affiliation: | 1. State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing, 100876, PR China;2. Web Sciences Center, School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu, 611731, PR China;3. Big Data Research Center, University of Electronic Science and Technology of China, Chengdu, 611731, PR China;1. BITS Pilani-Goa Campus, Zuarinagar, India;2. Department of Electrical & Computer Engineering, Ryerson University, Toronto, Ontario, Canada;1. School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China;2. School of Computer Science and Information Engineering, Harbin Normal University, Harbin 150025, China |
| |
| Abstract: | This paper proposes the application of Covariance Matrix Adaptation Evolution Strategy (CMA-ES) in fixed structure H∞ loop shaping controller design. Integral Time Absolute Error (ITAE) performance requirement is incorporated as a constraint with an objective of maximization of stability margin in the fixed structure H∞ loop shaping controller design problem. Pneumatic servo system, separating tower process and F18 fighter aircraft system are considered as test systems. The CMA-ES designed fixed structure H∞ loop-shaping controller is compared with the traditional H∞ loop shaping controller, non-smooth optimization and Heuristic Kalman Algorithm (HKA) based fixed structure H∞ loop shaping controllers in terms of stability margin. 20% perturbation in the nominal plant is used to validate the robustness of the CMA-ES designed H∞ loop shaping controller. The effect of Finite Word Length (FWL) is considered to show the implementation difficulties of controller in digital processors. Simulation results demonstrated that CMA-ES based fixed structure H∞ loop shaping controller is suitable for real time implementation with good robust stability and performance. |
| |
| Keywords: | Evolutionary algorithm Covariance matrix adaptation evolution strategy PI controller |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
