基于神经网络的转子系统故障诊断

Fault Diagnosis of Rotor System Based on Kohonen Neural Network

提出了一种基于神经网络的转子振动故障诊断的新方法,该方法以大型机器的轴承振动裂度作为神经网络的训练样本输入,并通过神经网络的学习、聚类,产生神经网络聚类中心,根据网络聚类的特点以及聚类的中心来判断转子的振动特性和实质。实例验证表明,该方法可实现对转子系统振动故障的准确诊断。

This paper puts forward a novel method for fault diagnosis of rotor system based on Kohonen neural network. The values of vibration crack degree of bearings in large machines are input to Kohonen network as training sample,and clustered by the network. To diagnose vibration faults of rotor system ,the vibration characteristics and substances of rotor are determined according to clustering center. A real example shows that Kohonen neural network is a meritorious fault diagnosis means of rotor system.