Improvement And Research On Fault Diagnosis Method Of Rolling Bearing

In recent years,with the continuous advancement of science and technology and the continuous development of industry,intelligent devices are gradually tending towards automation and intelligence.At the same time,urban rail transit equipment is becoming increasingly complex and urban traffic jams are common.Therefore,accurate and timely fault diagnosis of equipment can reduce its possible serious consequences effectively,save costs and reduce risks.The fault of equipment mainly comes from rolling bearings,but the original fault diagnosis technology can not meet the high requirements of modern intelligent equipment for fault diagnosis totally.Therefore,the fault diagnosis of rolling bearing is taken as the background in this thesis,diagnoses and studies the three types of faults of inner ring fault,outer ring fault and roller fault.In order to improve the efficiency and accuracy of rolling bearing fault diagnosis,and combining the research status of rolling bearings at domestic and overseas,a new type of fault diagnosis method is presented in this thesis.Firstly,the vibration signals under different fault conditions are collected.Secondly,wavelet packet decomposition and reconstruction are used to extract the energy feature vector that can reflect different fault states in the vibration signals.Then,genetic algorithm is used to optimize BP neural network(GABP),Biogeography-based optimization algorithm is used to optimize radial basis neural network(BBO-RBFNN),and GABP diagnosis model and BBO-RBFNN diagnosis model are obtained.Finally,the extracted energy feature vector is input into the GABP model and the BBO-RBFNN model for fault identification and diagnosis.At the same time,MATLAB software is used to carry out simulation analysis of the collected data.Through simulation experiments,it is verified that compared with the unoptimized neural network,the optimized neural network model has higher accuracy in diagnosis of faults,and the convergence speed is faster and has a good diagnostic effect.