Research And Implementation Of Route-Planning Algorithm For Transportation Network Based On Cloud Computing

In recent years, heavy traffic and road incidents has brought great influence on people’s dailylives, especially in those metropolis. Intelligent Transportation System is considered the best approachto solve this problem. The Route Guidance System is an important component of IntelligentTransportation System, while the key of Route Guidance System is route guidance algorithm. Thus, itis crucial to improve the efficiency of the route guidance algorithm. With the constant development ofeconomy and urbanization process, cities have been expanding as well as the data scale of transportsystem. In this paper, for the demand of mass data environment and high-performance processing, wedeveloped the distributed parallel algorithm target at improving the system’s efficiency, which isbased on the Intelligent Transportation cloud-based platform. The main contents of this dissertationare expressed as follows:Firstly, the core technologies of cloud computing have been analyzed, especially the distributedfile system(GFS), distributed computing programming model-MapReduce, and distributed datamanagement technology(Bigtable). And the focuses are hadoop Distributed File System and hadoopdistributed database.Secondly, route guidance algorithms and their characteristics are briefly introduced. Upon this,Ant colony algorithm is improved and researched on parallelization. And under hadoop, the algorithmbased on the programming model of MapReduce has been designed and implemented.Thirdly, on the basis of MapReduce and the management technology of distributed data, themanagement method of the transportation network data is researched. By using of traffic simulationsoftware VISSIM, transportation data is analyzed and the data structure of transportation network isdesigned. Through the establishing of the abstract data model of the transportation network for the antcolony algorithm, the parallel ant colony algorithm based on cloud computing is implemented.Finally, a platform of cloud computing hadoop is built. And parallel ant colony algorithm basedon cloud computing is experimented and analyzed. The efficiency for the search of shortest path hasbeen improved.