Research On Fault Diagnosis Method Of Rolling Bearing Based On VMD-HHT Marginal Spectrum And SSA-DBN Model

Rolling bearing plays an extremely important role in modern life.If it breaks down,it will have a great impact on daily life and inevitably cause economic loss.Because of the variety and uncertainty of rolling bearing faults,it is difficult to determine its state.Therefore,the fault diagnosis classification of rolling bearing vibration signals is of great engineering significance and application value.This paper presents a fault diagnosis method based on VMD-HHT marginal spectrum method and SSA-DBN network model from two aspects of fault feature extraction and fault diagnosis model of rolling bearing:(1)It is found that the vibration signal of rolling bearing is non-linear and non-stationary by analyzing the current fault diagnosis technology.In this paper,the method of combining Variational Mode Decomposition with Hilbert-Huang Transform marginal spectrum to form VMD-HHT marginal spectrum is proposed to extract the fault characteristic components of rolling bearing vibration signals.Experimental results show that this method has a good performance in the feature extraction of rolling bearing vibration signals.(2)In order to reduce the influence of random factors and improve the accuracy of diagnosis,the SSA-DBN model is composed of Deep Belief Network and Sparrow Search Algorithm,and the weight parameters are optimized by SSA.Because the formula of SSA is not strict,the SSA is modified according to the references,and the feasibility of the modified SSA is verified by comparing Particle Swarm Optimization,Artificial Bee Colony Algorithm and Immune Algorithm.Experiments with UCI common data sets show that SSA-DBN model can improve the diagnostic accuracy compared with DBN model.(3)Based on the rolling bearing data set of Case Western Reserve University,the VMD-HHT marginal spectrum method is used to extract the fault feature vectors and input them into DBN model and SSA-DBN model respectively for fault diagnosis.The experiment shows that VMD-HHT marginal spectrum method can be used in combination with DBN model and SSA-DBN model,the SSA-DBN model has higher diagnostic accuracy than DBN model.The method proposed in this paper has the ability to accurately identify the classified states from the various bearing operating states,and has a high accuracy,has a certain application value.At the same time,it can also provide a reference for the diagnosis of other non-linear and non-stationary equipment.