基于SOM神经网络的风机故障诊断研究

Study on the Fault Diagnosis of Fan Based on SOM Neural Network

应用自组织特征映射(SOM)神经网络实现风机的故障诊断。介绍了SOM网络的结构和学习算法;总结了风机的故障集、征兆集和故障特征数据。在MATLAB环境下给出了风机故障诊断的具体实例,表明该方法是一种可行有效的风机故障诊断方法。

In this paper, a method based on self-organizing featureMap neural network is applied to realize fault diagnosis of fan. The structure of SOM network and learning algorithm are introduced. The fault sets, fault symptoms and fault feature data are summarized. An instance of fault diagnosis of fan is presented in Marlab environment, which indicates this method can effectively diagnose the fault of fan.