摘要
针对永磁直线同步电动机的端部效应和非线性摩擦问题,采用一种鲁棒自适应神经网络控制方法,实现了永磁直线电机的跟踪控制。所设计的控制器包含两个部分:一部分是自适应神经网络控制器,用来逼近理想控制器,该神经网络的输入为滑模切换函数;另一部分是鲁棒控制器,用来消除逼近误差。通过李亚普诺夫稳定性定理验证了所设计的控制器能够保证控制系统渐进稳定。仿真结果表明:所设计的控制器能达到较好的控制效果。
For the problems of the end effect and the nonlinear friction of servo system of permanent magnet linear synchronous motor,a robust adaptive neural network control scheme was adopted,and this scheme realizes the position tracking performance of the servo system. The robust adaptive neural network control scheme is comprised of a neural network controller which is utilized to approximate the perfect controller and a robust controller which is used to eliminate the approximate error. The input of the adaptive neural network is sliding switching function. The designed controller insures the asymptotical stability of the control system, what is certificated by Lyapunov stability theorem. The simulation results demonstrate the scheme has better control precision.
出处
《电气传动》
北大核心
2011年第11期54-58,共5页
Electric Drive
基金
国家自然科学基金项目(60964001)
广西重点自然科学基金(桂科自0991019Z)
广西信息与通讯重点实验室基金项目(10902)
关键词
鲁棒自适应神经网络
永磁直线同步电机
端部效应
摩擦
robust adaptive neural network control
permanent magnet linear synchronous motor (PMLSM)
end effect
friction