Research on Roller Bearing with Fault Diagnosis Method Based on EMD and BP Neural Network

• DOI: 10.4028/www.scientific.net/amr.1014.501
• Publication Date: 2014-07-29T08:51:33Z
• Authors: Shu Guo; You Cai Xu; Xin Shi Li; Ran Tao; Kun Li; Min Gou
• Keywords: Kurtosis; SIGNAL (programming language); Feature vector; Feature (linguistics)

In order to discover the fault with roller bearing in time, a new diagnosis method based on Empirical mode decomposition (EMD) and BP neural network is put forward paper. First, we get signal through experiments. Then use EMD decompose vibration into series of single signals. We can extract main information from The kurtosis coefficient signals forms feature vector which used as input data network. trained be for identification. Through analyzing, distinguish normal state, inner race fault, outer fault. results show that this gain very stable classification performance good computational efficiency.