直线永磁同步伺服系统的滑模-神经网络控制

Sliding Mode-Neural Network Control for PMLSM

针对直接驱动的直线永磁同步伺服系统,提出一种基于滑模控制和神经网络控制相结合的双自由度控制策略。滑模输入控制器保证了系统对给定的快速跟踪性能;神经网络输出反馈控制器对系统参数摄动和外在阻力变化进行抑制,并削弱了滑模控制引起的系统抖振。该控制策略很好地解决了直线永磁同步伺服系统的跟踪性能和鲁棒性能之间的矛盾。仿真结果表明该方案在保证伺服系统快速性的同时,对参数摄动和阻力扰动(尤其是非线性时变扰动)具有很好的鲁棒性。

A two degree of freedom control scheme on combing sliding mode control and neural network control for the permanent magnetism linear synchronous motors （PMLSM） ：the input controller was designed on sliding mode control method, which can ensure the system had fast tracking characteristic to the command ;the output feedback controller was realized by a neural network, which suppressed parameter variations and resistance perturbation, and eliminated the chattering of sliding mode control. This control scheme solved the conflict between tracking and robustness of the system. The simulation results showed that this control scheme not only enhanced the track -command -ability, but also had strong robustness to parameter variations and resistance perturbation （ especially the nonlinear and time - varying perturbation）.