Study On The Relativity Between Degradation State Of Wagon Bogie And 4T Monitoring Data

With the transformation of the railway truck maintenance method from Maintenance by plan to Maintenance by status,it is of great guiding significance for realizing and guiding Maintenance by status to accurately judge the status of key components of railway truck.The current railway truck safety monitoring system is mainly used to monitor the real-time status of railway wagons and maintain the daily safety of railway trucks.With the online monitoring information collected in the actual operation of the truck and the inspection information of the railway truck,we are able to extract the characteristics related to the degradation of the components in the monitoring data,explore the correlation between the online monitoring information and the degradation state and improve the utilization of the online monitoring information,as well as to achieve timely pre-judgment of component status and maintenance guidance,improve system reliability and reduce maintenance costs.In this thesis,we take the bogie as our study object and analyze the relationship between its degeneration and the information in the carriage safety operation monitoring system,or simply 4T System.Firstly,to analyze the bogie’s maintenance information,we take wheel,bearing,bolster and side frame as the key parts reflecting the bogie’s degeneration status,Then,the maintenance information reflecting the degradation state of each part could be extracted.And the state of bogie is judged by K-Means clustering,and the clustering analysis results are verified by fuzzy comprehensive evaluation,thus more accurate state evaluation results are obtained.In the meantime,we extract the information in 4T System and establish a characteristics set.Subsequently,normalized processing is carried out based on the characteristics.Then,using the monitoring information and bogie status information,the process of classification supervision is constructed.In this thesis,five representative algorithms,which are Random Forest(RF),Extremely randomized trees(ERT),Gradient Boosting Decision Tree(GBDT),Support Vector Machine(SVM)and Logical Regression,are used to learn the classification problem,and the corresponding classification algorithm model is established.The case proves that there is a correlation between the degraded state of railway wagons and the "4T" information,and the tree structure algorithm is more suitable for the study of the relationship between them。The results of this study provide a certain basis for the upgrading of railway vehicle information system,the health evaluation of spare parts and the formulation of maintenance strategy,and help the railway freight car industry develop towards high reliability.