Optimization-based adaptive neural sliding mode control for nonlinear systems with fast and accurate response under state and input constraints |
| |
| Institution: | 1. State Key Laboratory of Fluid Power and Mechatronic Systems, Zhejiang University, Hangzhou, 310027, People’s Republic of China;2. Ocean College, Zhejiang University, Zhoushan, 316021, People’s Republic of China;3. Ocean Research Center of Zhoushan, Zhejiang University, Zhoushan, 316021, People’s Republic of China;4. Hainan Institute of Zhejiang University, Sanya, 572025, People''s Republic of China;1. School of Mechanical Engineering, Yanshan University, Qinhuangdao 066004, China;2. Institute of Complexity Science, Qingdao University, Qingdao 266071, China;1. School of Engineering, Huzhou University, Huzhou 313000, China;2. School of Science, Huzhou University, Huzhou 313000, China;3. School of Science and Technology, Huzhou College, Huzhou 313000, China;4. School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China;5. School of Physics and Electronic Information, Anhui Normal University, Wuhu 241000, China;1. HSE University, Moscow, Russian Federation;2. School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, USA;3. CINVESTAV-IPN, Department of Control Automatic, UMI 3575 CINVESTAV-CNRS, Mexico City, Mexico;4. Institut für Theoretische Informatik, Mathematik und Operations Research, Universität der Bundeswehr München, München, Germany;1. College of Mathematics and Computer Science, Tongling University, Tongling, 244000, China;2. School of Electrical and Information Engineering, Jiangsu University, Zhenjiang, 212013, China;1. Facultad de ingenieria Mecanica y Electrica, Universidad Autónoma de Nuevo León, San Nicolas De Los Garza, Nuevo, 66451 León, México;2. Nantes Uiniversité, Ecole Centrale de Nantes, LS2N UMR CNRS 6004, Nantes, France |
| |
| Abstract: | The high-performance control requires the system to be stable, fast and accurate simultaneously. However, various systems (e.g., motors, industrial robots) generally face technical challenges such as nonlinearities, uncertainties, external disturbances and physical constraints, which make it difficult to reach the hardware potential of the systems to track the desired trajectories when satisfying the high-performance control requirements. Therefore, take a two-order nonlinear system for example, an optimization-based adaptive neural sliding mode control based on a two-loop control structure is proposed in this paper, where the outer and inner loops are designed separately to achieve different control requirements. Namely, the outer loop is designed as a model predictive control (MPC)-based optimization problem, which can optimize the desired trajectories to meet the state and input constraints, and maximize the converging speed of transient response as fast as possible, and the inner loop is designed with a recurrent neural network (RNN)-based adaptive neural sliding mode controller, which can guarantee the tracking of the replanned desired trajectories from outer loop as accurate as possible. The stability of the system is guaranteed by Lyapunov theorem, the optimal tracking performance is achieved under nonlinearities, uncertainties, external disturbances and physical constraints, and comparative simulation with a motor system is carried out to verify the effectiveness and superiority of the proposed approach. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
