永磁同步电机变结构模糊神经网络控制策略  

Deformable Fuzzy Neural Network Control Algorithm of PMSM

作　　者：梁国伟 康忠健[1] LIANG Guowei;KANG Zhongjian(Colllege of New Energy,China University of Petroleum(East China),Qingdao 266580,China)

机构地区：[1]中国石油大学(华东)新能源学院,青岛266580

出　　处：《组合机床与自动化加工技术》2024年第7期83-88,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家重大科技专项项目(2016ZX05034004)。

摘　　要：为改善永磁同步电机矢量控制技术在复杂工况下控制器参数不能做出实时调整导致控制性能差的问题,分析了模糊逻辑与神经网络控制原理,提出了一种基于高斯径向基神经网络与模糊控制的相结合的智能控制策略。以转速误差以及误差的变化率为依据构建增量补偿式二维变结构模糊神经网络PID控制器(deformable fuzzy neural network,DFNN)通过RBF神经网络参数辨识器获取永磁同步电机的雅可比信息矩阵(Jacobian matrix),通过变结构算法确定变结构模糊神经网络的结构信息。在MATLAB/Simulink中仿真结果表明,该控制系统提升了电机启动以及目标转速发生改变时的响应速度,同时降低了超调量,在负载转矩存在扰动时转速变化小,且能够快速回归至给定值,优化了矢量控制系统的性能。To enhance the deficiencies of conventional vector control technology of permanent magnet synchronous motor,which pertains to the incapacity to dynamically modify controller parameters when the motor work in complex conditions,resulting in subpar control system performance.A combined intelligent control strategy based on Gaussian radial basis function neural network and fuzzy control has been proposed.Based on the speed error and the rate of change of the error,a incremental compensation-type two-dimensional deformable fuzzy neural network PID controller is constructed.The Jacobian matrix of the permanent magnet synchronous motor isobtained through the RBF neural network parameter identifier.The structure information of the deformable fuzzy neural network is determined through the variable structure algorithm.Simulation results in MATLAB/Simulink show that the control system improves the response speed and reduces overshoot during motor startup and when the target velocity changes.The velocity alteration appears negligible upon encountering an abrupt alteration in the load conditions,promptly reverting to the prescribed value and thus optimizing the functionality of the vector control system.

关 键 词：永磁同步电机 矢量控制技术 智能控制 变结构模糊神经网络 

分 类 号：TH165[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]