Research On Technologies Of Fault Warning For Automobile Assemble Conveying Equipment

With the development of the intelligent system,automobile assembly transport equipment tends to be sophisticated and multifunctional,causing the difficulty with daily maintenance.In order to solve the problem with low efficiency and high cost in the maintenance and meet the demand of real-time fault early warning as well as visualization,the paper makes the research on fault early warning technologies of automobile assembly transport equipment.The research is conducted from three aspects: fault diagnosis,state evaluation and information visualization.The paper proposes the original methods and realizes the fault early warning system development for automobile assembly transport equipment based on Web framework.Firstly,the paper presents a fault diagnosis framework and a new method for the global equipment through signal processing and machine learning algorithms.Empirical wavelet transform(EWT)and singular value decomposition(SVD)are used to extract signal feature.Improved multi-classified Adaboost algorithm based on multi base classifier is utilized to implement fault classification.Finally,the bearing,a usual part in automobile assembly transport equipment,is used to verify the algorithm.On the field of state evaluation,the paper proposes a state assessment of mechanical equipment based on information entropy and SOM neural network.An outlier detection method based on density is used to remove noise points and detect the abrupt fault.Information entropy method is introduced in order to give variation impact factor for different assessment attributes.Based on these,a new approach is used to assess health state by integrating with variation impact factor and SOM neural network.Then,the belt elevator is utilized to verify the results of the proposed state assessment method.Based on the research on fault diagnosis and state evaluation,the paper further analyzes the information visualization for fault early warning and shows the diagnosis consequences by visualization.Meanwhile,the paper introduce the association rules mining for fault early warning and realize it through Apriori algorithm.Finally,fault early warning system development framework for automobile assembly transport equipment is put forward,which is validated by a real application of company.