Fault diagnosis approach to vehicle hub bearing unit based on danger theory

In order to treat the strong nonlinear characteristics of noise and vibration signals during the fault diagnosis from the vehicle hub bearing unit, this work proposes a novel fault diagnosis algorithm of the unit based on immune Danger Theory. The wavelet packet technology is used in the proposed algorithm to extract multiple characteristic parameters such as standard deviation, energy, variance, and kurtosis. Based on them, a characteristic cell array vector is constructed to reflect the real operating status of the hub bearing unit. According to the Danger Theory model, three kinds of signals and the danger interval are defined. The vibration signals of the Type 606 vehicle hub bearing collected from the bearing test bench are used to verify the effectiveness of the proposed algorithm. The experimental results show that the operating status of a hub bearing unit in the different working conditions can be detected by the proposed algorithm.