基于径向基函数神经网络的无刷直流电机多机电故障处理方法  

Treatment Methods for the Multiple Electromechanical Faults of Brushless DC Motors Based on the Radial Basis Function Neural Network

作　　者：常玉燕 CHANG Yuyan(Suzhou Vocational and Technical College,Suzhou,Anhui Province,234001 China)

机构地区：[1]宿州职业技术学院,安徽宿州234001

出　　处：《科技资讯》2024年第12期66-70,77,共6页Science & Technology Information

基　　金：安徽省高等学校省级质量工程项目“新能源汽车技术专业知名工匠培养基地”(项目编号:2022zmgj042)。

摘　　要：解决无刷直流电动机的故障检测和分类问题,提出了一种新的诊断方法,可用于定位定子间转、转子动力和静力不平衡等多种机电故障。结合无刷直流电动机电流信号、电机转矩和速度信息,利用小波包变换提取故障特征,并将其作为径向基函数神经网络的输入数据。通过粒子群优化算法和遗传算法对神经网络权值进行更新,提高了算法的效率和灵活性。最终,通过比较不同神经网络和优化方法的组合结果验证了该方法的有效性。In order to solve the problem of the fault detection and classification of brushless DC motors,this paper proposes a new diagnostic method that can be used to locate various electromechanical faults such as the inter-turn fault of the stator,and the dynamic and static imbalance of the rotor.Combined with the current signal,motor torque and speed information of the brushless DC motor,it extracts fault features by the wavelet packet transform and uses them as the input data of radial basis function neural networks.By the particle swarm optimization algo-rithm and genetic algorithm,it updates the weights of the neural network,which improves the efficiency and flex-ibility of algorithms.Finally,it verifies the effectiveness of the proposed method by comparing the combined results of different neural networks and optimization methods.

关 键 词：无刷直流电动机 小波包变换 神经网络 粒子群优化算法 遗传算法 

分 类 号：TM33[电气工程—电机]