基于GA优化RBF网络的永磁同步电机无位置控制  

Sensorless Control of Permanent Magnet Synchronous Motor Based on RBF Neural Network With Genetic Algorithm Optimization

作　　者：金昊 丁曙光[1] JIN Hao;DING Shu-guang(School of Mechanical Engineering, Hefei University of Technology, Hefei 230009, China)

机构地区：[1]合肥工业大学机械工程学院,合肥230009

出　　处：《组合机床与自动化加工技术》2021年第2期95-98,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：国家自然科学基金资助项目(51775156);安徽省科技厅秋实计划资助项目(2013AKKG0392)。

摘　　要：在表贴式永磁同步电机(surface permanent magnet synchronous motor,SPMSM)的无位置控制中,准确实时地估计其转子位置和转速是一个关键技术难点。文章把遗传算法(genetic algorithm,GA)和径向基函数(radial basis function,RBF)神经网络相结合,提出了GA-RBF神经网络观测器来估计SPMSM的转子位置和转速。GA-RBF观测器首先估计了两相静止坐标系下的反电动势,然后引入锁相环实时估计了转子位置和转速。仿真结果表明,GA-RBF算法能够提高神经网络的收敛性,避免神经网络陷入局部最优,从而实现对SPMSM的高精度控制。In the sensorless control of surface permanent magnet synchronous motor(SPMSM),it is a key technical difficulty to accurately estimate the rotor position and speed in real time.Combining genetic algorithm(GA)and radial basis function(RBF)neural network,a GA-RBF neural network observer is proposed to estimate the rotor position and speed of SPMSM.The GA-RBF observer first estimates the back electromotive force in two-phase static coordinate system,and then introduces a phase-locked loop to estimate the rotor position and speed in real time.The simulation results show that GA-RBF algorithm can improve the convergence of neural network and avoid neural network falling into local optimum,so as to achieve high-precision control of SPMSM.

关 键 词：永磁同步电机 径向基函数神经网络 遗传算法 无位置控制 

分 类 号：TH16[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]