Fault Diagnosis Modeling And Key Technology Research For Traffic Vehicle Rolling Bearing

As the direct carrier of delivering passengers in transportation field, it has always been the focused issue for the reliability and security of the transportation vehicle.Rolling bearing is the key component of transportation vehicle, in the condition of high rotating speed and loading, the rolling bearing is apt to be abraded and become faulted,certain traffic damage will happen if not replaced timely. Therefore, we should research on fault diagnosis key technology and improve accuracy, the rate of traffic accident will reduce and the total safety level will be improved.Taking rolling bearing as the research object in this dissertation which is the typical rotating component in traffic vehicle, this author researched on the fault diagnosis critical technology. The main content and innovations are as follows:(1) The typical structure and fault mode for rolling bearing were deeply studied.Vibration signal was chosen as the running state monitoring signal for rolling bearing.And then, the vibration mechanism and typical fault mode characteristic for rolling bearing were expounded.(2)In allusion to vibration signal filtering for rolling bearing. A signal filtering method named 'Auto-adapted Morphological D-value Filter' was proposed. The noise and low frequency interference is able to be restrained effectively by this method.(3)Aiming at the feature extraction issue for rolling bearing,a method based on Morphological Entropy was proposed taking multi-scale mathematical morphology and complexity theory. Different operation conditions and fault modes are able to be reflected quantitatively by Morphological Entropy.(4)As to the fault diagnosis issue for rolling bearing, a fault diagnosis model based on SOA-multi-classification SVM diagnosis model was proposed. In this method,optimization for multi-classification SVM diagnosis model is based on seeker optimization algorithm which has a fast convergence rate and high integrated efficiency.(5)On the basis of fault diagnosis key technique analysis, a flow of fault diagnosis for rolling bearing is constructed. The vibration signal of rolling bearing is the analysis object, the signal preprocessing method is Auto-adapted Morphological D-value Filter,the fault mode feature vector is Morphological Entropy of vibration signal multi-classification SVM model is chosen as fault diagnosis model and SOA is proposed in parameters optimization,establishing an optimal multi-classification SVM model, realizing the recognition of different fault mode for rolling bearing.To verify the effectiveness of the method, simulation signal and rolling bearing instance data from the Case Western Reserve University is introduced in the validation,the results showed that the flow proposed is able to realize effective diagnosis of rolling bearing in four different states in four different load conditions.