High Speed Train Axle Temperature Data Analysis System Towards Real Time Fault Diagnosis

With the rapid development of China’s transportation industry,the running speed of high-speed trains continues to increase that leads to higher safety requirements.As a key point to ensure the safety,fault diagnosis of train operation is discssed in this paper by analyzing the data of high-speed trains in real time in order to locate the faults of high-speed trains and avoid or reduce the accidents.Supported by the National Natural Science Foundation of China’s major project "high speed train information control system fault modeling theory and method based on big data and knowledge",this paper puts forward a real-time fault diagnosis of high-speed train shaft temperature data analysis system.The major contributions are as follows:(1)The existing fault diagnosis systems are reviewed,and the status of the fault diagnosis system and big data analysis platform architecture of high-speed trains is investigated.Because of the big data characteristics of high-speed train,i.e.,large amount of data,variety,low value density and high real-time performance,high speed train fault diagnosis system not only needs to support large-scale data analysis but also real-time requirements.Due to the lack of an analysis system for real-time fault diagnosis of large-scale high-speed train data,a high-speed train bearing temperature data analysis system is proposed in this paper.The functional requirements of the system is analyzed in detail from the perspective of the data flow of high-speed trains and the requirements of the system from three aspects:data analysis requirements,data transmission requirements,and data storage requirements.(2)In accordance with the principle of modular design,the design of high speed train bearing temperature data analysis system architecture for real-time fault diagnosis is designed.The system mainly includes the following functional modules:data acquisition module,data transmission module,data preprocessing module,data storage module,data analysis module and data visualization module.The data acquisition tool is used to load the high-speed train data into the data transmission module to achieve real-time data acquisition.By constructing the message queue to realize real-time data transmission,the real-time problem of high-speed train data transmission is solved.The preprocessing tool can cleanse the high-speed train data and solve the problem of the high-speed train data format.By proposing a lossless compression method to efficiently compress the high-speed train data,the storage space is greatly reduced.The large-scale data occupancy of the high-Space,the storage of large-scale high-speed train data is realized by constructing distributed database,the real-time analysis of high-speed train data is solved by deploying real-time data analysis engine.The visualization of high-speed train data is solved by data visualization tools.(3)Based on the industrial cloud platform of State Key Laboratory of Process Industry Automation,the system has been developed using Java development language,MySQL database,and big data processing framework such as Hadoop,Zookeeper,Rabbitmq,Storm,HBase and visualization software Tableau Desktop Development.Among them,Hadoop is used to build the infrastructure.The main use of its distributed file system HDFS to store data files generated by high-speed train to solve the problem of large-scale data files generated during the operation of high-speed trains;Zookeeper used to do resource coordination,At the same time,it stores some metadata information of Storm and Rabbitmq to realize the reliability of the analysis system.Rabbitmq is used to transmit real-time data of high-speed train to real-time problem of high-speed train data transmission.Storm is used to solve the real-time problems of high-speed train data analysis;HBase is used to store high-speed train historical data and compressed historical data to store the large-scale high-speed train historical data;MySQL is used to store high-speed train real-time results from data and data analysis modules,as a data source for data visualization modules,reduce the complexity of visualizing high-speed train data;Tableau Desktop is used to visualize data and enable visualization of high-speed train data that makes it even better.(4)The system is deployed in Industrial Cloud Computing Center of State Key Laboratory of Process Industry Automation.A number of functional modules of high-speed train shaft temperature data analysis system for real-time fault diagnosis are experimentally verified.The experimental results show that the system of real-time fault diagnosis for high-speed train shaft temperature data proposed in this paper realizes real-time data acquisition,real-time transmission,large-scale data storage,high speed train data visualization,and real-time warning and alarm.The experimental results show that the system meets the real-time requirements well..