Research on Vehicle Fault Diagnosis Performance of GBDBN Model

Vehicle traditional fault diagnosis technology consumes a lot of manpower, it is difficult to excavate the internal rules of fault signals and to guarantee the accuracy of diagnosis results for a large number of data. Deep learning method has been widely used in fault diagnosis. This paper proposes a new method based on Gaussian-Bernoulli deep belief network. Aiming at the problem that there is no theoretical basis for parameter setting in the process of model training, taking the vibration acceleration signal of rolling bearing as the research object, the applicability of some parameters is experimentally studied, and the influence of different parameters on the fault diagnosis performance of GBDBN model is discussed. To some extent, the experimental results provide a reference for setting the initial parameters of GBDBN model in the training process.