Application And Research On Data Storage Of Rail Transit Maintenance Support System Based On Hadoop

With the rapid development of urban rail transit, rail transit safety is paid more and more attention, which will certainly make higher requirements for maintenance and management. Maintenance Support System has emerged and become the focus of research and application. In the process of maintenance and management, the data of real-time monitoring signal system is growing, which is far beyond the traditional database storage scope, the methods of traditional data storage and data processing has encountered a bottleneck. Cloud computing Hadoop provides a new way to solve the above problems as a new model, which has characteristics of large scale, high scalability and high reliability.Firstly, the importance of the Maintenance Support System of rail transit is explained in the introduction, and MSS brings value-added services combined with Big Data. Secondly, the research situation of Maintenance Support System and the research situation of data storage technology and Hadoop are summarized and concluded carefully. The features and structure of MSS are analyzed, and comparing the advantages and disadvantages of traditional software and Hadoop technology, the current problems in the data storage and query are further defined. Therefore, Hadoop technology is applied into MSS for reliable storage, efficient query and data mining analysis.For different types of MSS data, combining Hadoop cluster technology, as well as the strength of the hardware and software, MSS data storage design based on Hadoop technology is proposed in this paper. MSS data storage is studied deeply from the whole storage architecture, Hadoop distributed file system and HBase distributed database in the article. Firstly, a log collection system Flume is used to transmit data and the storage ways of MSS data in HDFS is concluded. Secondly, Storage module of maintenance data based on HDFS is designed in this article, and how to read and write data in HDFS is described by giving an example. Lastly, the data storage in HDFS is realized. According to the characteristics of MSS data, HBase storage model of MSS data element is designed. Logical and physical models of HBase data are established by taking ZC data for example, and data is written to the HBase database successfully.Depending on the theory, Hadoop platform environment which is fully distributed module is built, including hardware and software environment. It is feasible that Hadoop technology is applied to Maintenance Support System of rail transit by using system performance analysis and testing storage performance of Hadoop clusters. In addition, Hadoop has the advantage over the traditional way in the aspect of data storage, and the data is larger, the advantages of Hadoop are more obvious.In the end, the work is summarized and the innovation is described, as well as the deficiencies and areas where improvement is needed are proposed, and research direction for future work is discussed in this paper.