Fault dignosis of rolling bearing based on time domain parameters

Abstract

The rolling bearing is the common component in machinery. Its running state will influence the performance of the whole machine directly. In this paper we put forward a feature extraction method of fault diagnosis of rolling bearing. After the vibration signals of the rolling bearing are analysed and processed, the feature parameters which represent operating state of the rolling bearing are extracted, and then are inputted to the BP neural network to train the network with BP algorithm by processing of normalization. Good rolling bearings and bad rolling bearings can be identified with this network. The simulation result shows that the method presented in this paper is practical and effective.