Study Of The Model And Algorithm For The Acquisition And Prediction Of Railway Wagon Flow Path Based On Hadoop Platform

In recent years,China Railway Company actively implements the reform of railway freight transportation,tries the best to improve the level of railway freight transportation services,strengthens the construction of the railway logistics,and strives to expand the share of railway transportation in the market of freight transportation.But now,the most prominent problem in current railway transportation is the uncertainty of transportation time for goods,which restricts the competitiveness of the railway transportation.Although the railway department has formulated the time limit of freight transport,the freight transportation time will also be affected by the transport operations such as traffic rally,traffic transportation operations and so on,so railway department needs to strengthen the effective control of wagon flow to improve the efficiency of transport organization.The accurate prediction of wagon flow can achieve the effective control of wagon flow,avoid the occurrence of wagon flow congestion in time and ensure the transportation efficiency of railway network.Wagon flow path is the basis of wagon flow prediction and the confirmation of reasonable wagon flow path is very important for wagon flow prediction and transportation organization.This paper aims at the configuration of reasonable wagon flow path,uses the integration data of all business systems provided by the Railway Transportation Information Integration Platform,builds the wagon flow operation path acquisition model to obtain wagon flow path in the real situation and uses the Hadoop platform to handle wagon flow operation path,builds the patterns of wagon flow operation path and probability suffix tree according to different goods type in each station and verifies it in the wagon flow forecast system.The specific work are as follows:(1)This paper studies the wagon flow operation path reflecting the wagon flow path in real situation,obtains the integrated and shared business data of railway from Railway Transportation Information Integration Platform and builds the node mapping model、wagon report matching model,path sequence joint model to get the complete wagon flow path.(2)Considering the amount of railway wagon flow path data will be accumulated over time increasing,traditional data analysis methods can not accomplish effective analysis of the huge amounts of data,so this paper uses big data analysis method to process the wagon flow path data by Hadoop platform,uses Sqoop tools to implement the data transmission between traditional relational database and Hadoop Distributed File System,uses MapReduce programming model to deal with the big data sets of wagon flow operation path,uses the variable order markov model to handle the wagon flow path sequences,establishes the patterns of the wagon flow path and builds probability suffix tree to predict the wagon flow path.(3)This paper implements the acquisition process of wagon flow operation path based on the Railway Transportation Information Integration Platform using Java programming,builds Hadoop platform and MapReduce programming to deal with the data set of wagon flow path,implements the extraction of wagon flow operation path patterns and the establishment of probability suffix trees,uses different type of goods and different path patterns to predict the wagon flow operation path and verifies it in the wagon flow forecasting system.