Low-speed performance research for permanent magnet synchronous linear motor based on nonparametric model learning adaptive control

• Main Authors: Cao Rongmin, Zhou Huixing, Hou Zhongsheng, Wu Yingnian
• Format: Conference Proceeding
• Language: English
• Subjects: Adaptation models; Artificial neural networks; Force; Friction; Mathematical model; Permanent magnet motors; Prediction algorithms
• DOI: 10.1109/ICEMS.2011.6073732

Based on influence of friction, permanent magnet synchronous linear motor (PMSLM) with contact surface frequently appears rough running and control precision deterioration during low-speed operation. The design of nonparametric model learning adaptive control(NMLAC) algorithm in linear motor system is studied, estimation of pseudo-partial-derivatives is discussed, controller is based directly on pseudo partial-derivatives derived on-line from the input and output information of PMSLM using recursive least squares type of identification algorithms. The simulation control results show that the algorithms exhibit such advantages as good robustness, PMSLM low-speed response, against exogenous disturbance and noise for time-varying systems with vaguely known dynamics, the proposed method can realize good online friction estimation and compensation, hence the control performance of PMSLM is improved and it outperforms traditional PID controller and neural networks(NN) control method.