The Application of LMD and BP neural network in gear fault diagnosis

For time-domain energy messages conveyed by vibration signals of different types of gear fault are different, a method based on local mean decomposition (LMD) and neural network is proposed to apply to gear fault diagnosis. The vibration signal of gear which is mainly considered as the research object in this paper is decomposed into a series of product functions (PF) by LMD method. Then a further analysis is performed by selecting the PF components which contain main fault information, the energy feature parameters of the selected PF components form a feature vector which is considered as the input data of the neural network. The BP neural network is trained and used for fault identification. Then gear fault diagnosis is ultimately realized. Through the analysis of gear with normal state, wear fault and teeth broken, the results show that LMD method can decompose the complex non-stationary signal into a number of PF components whose frequency is from high to low, the method has an adaptive feature. And the method based on LMD and neural network has a higher fault recognition rate than the method based on wavelet packet analysis and neural network.