概率神经网络在电机故障诊断中的应用

Application of Probabilistic Neural Network in Motor Fault Diagnosis

针对传统故障诊断方法的局限性,提出一种基于概率神经网络的诊断方法。以异步电机转子故障为例进行了故障诊断研究,通过选择故障样本来训练神经网络,将代表故障的信息输入训练好的神经网络后,由输出结果就可以判断发生的故障种类。仿真实验结果表明了基于概率神经网络的电机故障诊断方法的正确性。

For the limitations of the traditional fault diagnosis,a method based on probabilistic neural network diagnosis was proposed.The fault diagnosis study was done by choosing the induction motor rotor fault as an example and the neural network was trained by selecting the fault samples.The fault information was inputted into the trained neural network,and then the system could determine the incidence of failure types according to the output.The simulation results show that the method of the probabilistic neural network based motor fault diagnosis is feasible.