基于决策树与神经网络结合的滚动轴承故障诊断方法

his paper proposes an improved fault diagnosis method based on the fusion of decision tree and neural network. This method performs EMD decomposition on the rolling bearing vibration signal, uses a decision tree to predict the failure of the decomposition signal, and then uses the attribute fusion neural network to learn the decision tree prediction results, and integrates the fault features of the decision tree into the neural network classifier. The results show that this method has a higher fault recognition rate and can accurately and effectively identify the type of faults in rolling bearings.