一类基于人工神经网络的三相异步电机断相检测方法  

One method based on artificial neural networks for phase failure detection in three-phase asynchronous motor

作　　者：王景焕 WANG Jinghuan(College of Aeronautical Engineering,Nanjing Vocational University of Industry Technology,Nanjing 210000,China)

机构地区：[1]南京工业职业技术大学航空工程学院,江苏南京210000

出　　处：《电子设计工程》2023年第5期80-83,88,共5页Electronic Design Engineering

摘　　要：针对三相异步电机发生断相故障的现象,该文采用概率神经网络的方法,通过采样三相异步电机不同工作状态下的数据,训练一个人工神经网络,用训练后的人工神经网络对三相异步电机进行检测,通过观测人工神经网络的输出即可得知哪一相出现了断相故障。概率神经网络的特点是结构简单、训练速度快,特别适合模式分类。通过在SIMULINK环境下对三相异步电机进行模型仿真,随机抽取数据对人工神经网络进行训练和测试,验证出所设计的人工神经网络在判断三相异步电机断相故障时准确率达到80%以上。According to the phenomenon of broken phase failure in three-phase asynchronous motor,the article adopts a method of probability neural network by training an artificial neural network with different working state data,and uses the trained artificial neural network to detect which phase is broken by observing the outputs of the network.Probabilistic neural network is characterized by simple structure and fast training speed in artificial neural networks,which are especially suitable for pattern classification.By simulating the model of three-phase asynchronous motor in SIMULINK environment and testing the artificial neural network,we verified that the accuracy of the designed artificial neural network reached more than 80% when judging the broken phase fault of three-phase asynchronous motor.

关 键 词：三相异步电机 概率神经网络 人工神经网络 断相故障 

分 类 号：TN712[电子电信—电路与系统]