The Open Automation and Control Systems Journal

2015, 7 : 296-302
Published online 2015 April 17. DOI: 10.2174/1874444301507010296
Publisher ID: TOAUTOCJ-7-296

A Fault Diagnosis Approach for Urban Railway Gearbox Based on EMD Method and Elman Neural Network

Wang Xing , Qi Xiangdong , Li Baojin and Skopincev E. Vektorovich
School of Computer, Taiyuan University of Science and Technology, Shanxi, 030024, P.R. China.

ABSTRACT

As for the non-linearity and non-stationary characteristics of the vibration signals of urban railway gearbox, an efficient method for gearbox fault detection and diagnosis based on EMD (empirical mode decomposition) and Elman neural network is proposed. First of all, the original signals are decomposed into a number of IMFs (intrinsic mode function) by EMD. Secondly, the feature vectors are constructed. Finally, these eigenvectors as fault samples input to the Elman neural network. The recognition results show that the EMD and Elman neural network is effective in railway gearbox fault diagnosis. This approach can be used as a useful tool for the rotating machinery fault diagnosis.

Keywords:

Urban railway gearbox , EMD, Elman neural network, IMF , fault diagnosis .