基于MRAS的三相PMSM神经滑模控制  

MRAS-based neural sliding mode control over three-phase PMSM

作　　者：苗敬利[1] 秦王毓 王桂梅[1] MIAO Jingli;QIN Wangyu;WANG Guimei(Hebei University of Engineering,Handan 056038,China)

机构地区：[1]河北工程大学,河北邯郸056038

出　　处：《现代电子技术》2021年第7期133-136,共4页Modern Electronics Technique

基　　金：河北省自然科学基金重点项目(E2019402436)。

摘　　要：针对永磁同步电机启动慢、转速超调高、抖振强、观测误差大的特点,提出一种基于模型参考自适应系统的神经滑模控制策略。通过将径向基函数神经网络算法应用于滑模控制器中,对传统的滑模控制策略进行改进,实现对于转子位置和速度的快速有效追踪。同时,在控制器的设计中使用饱和函数取代传统切换函数,并采用指数趋近率,根据Popov超稳定性与正性动态系统理论设计模型参考自适应系统。通过进行Matlab仿真,结果表明所提算法具有可行性,并且能够实现永磁同步电机的快速稳定运行,具有良好的动静态性能。Since the permanent magnet synchronous motor(PMSM)has the deficiencies of slow start,large rotate speed overshoot,violent buffeting and large observation error,a neural sliding mode control strategy based on model reference adaptive system(MRAS)is proposed.The RBF(radial basis function)neural network algorithm is used for the sliding mode controller to improve the traditional sliding mode control strategy,so as to realize fast and effective tracking of the rotor position and speed.In the design of the controller,the saturation function is used to replace the traditional switching function,and the exponential approach law is adopted.The MRAS is designed on the basis of the Popov hyperstability theory and positive dynamic system theory.The results of Matlab simulation show that the proposed algorithm is feasible,can achieve the fast and stable operation of PMSM,and provide good dynamic and static performance for the system.

关 键 词：神经滑模控制 永磁同步电机 矢量控制 快速追踪 模型参考自适应系统 系统设计 MATLAB仿真 

分 类 号：TN876-34[电子电信—信息与通信工程] TM761[电气工程—电力系统及自动化]