Fault Prognosis and Diagnosis of an Automotive Rear Axle Gear Using a RBF-BP Neural Network

Abstract

The rear axle gear is one of the key parts of transmission system for automobiles. Its healthy state directly influences the security and reliability of the automotives. However, non-stationary and nonlinear characteristics of gear vibration due to load and speed fluctuations, makes it difficult to detect and diagnosis the faults from the transmission gear. To solve this problem a fault prognosis and diagnosis method based on a combination of radial basis function(RBF) and back-propagation (BP) neural networks is proposed in this paper. Firstly, a moving average pretreatment is used to suppress the time series fluctuation of vibration characteristic parameter tie series and reduce the interference of random noise. Then, the RBF network is applied to the pretreated parameter sequences for fault prognosis. Furthermore, based on self-learning ability of neural networks, characteristic parameters for different common faults are learned by a BP network. Then the trained BP neural network is utilized for fault diagnosis of the rear axle gear. The results show that the proposed method has a good performance in prognosing and diagnosing different faults from the rear axle gear.