Road Mapping And Analysis Application Based On Mobile Communication Positioning Data

With the rapid development of mobile communication technology and the widespread popularization of mobile terminal devices,an increasingly amount of space-time data of human beings is obtained,such as cell phone location data,cell phone application data and taxi GPS data.The data for these spatiotemporal marks contains user trajectory information for moving individuals.Obtaining the real moving trajectories of the users based on such data,mining the movement characteristics of the users and studying the traffic conditions of the roads have become key components of the intelligent transportation system.Among them,mobile location data becomes a hot topic in academic research because of its wide coverage,easy obtainment and other characteristics.Operators can analyze the transportation industry by mining the added value of these data and work with the transport authorities to rationalize transport capacity and public transport to plan the road network so as to ease the impact of traffic congestion on the lives of urban residents.Aiming at the demand of operators,the thesis designs and implements a distributed road traffic analysis system framework based on Spark platform,which realizes the functions such as data cleaning,base station data preprocess and trajectory segmentation of mobile positioning data.After obtaining the preprocessed user trajectories,we have implemented the functions of road mapping,congestion analysis,traffic mode determination and space-time query of trajectories.Aiming at the base station data,the thesis improves the traditional Hidden Markov Model by considering the complex traffic network,the location of the base station and driver's driving habits,and builds an improved map matching model for the complex traffic network.The algorithm can deal with the parallel roads and the road intersection situation effectively.Theoretical analysis and experimental test of this thesis evaluated the proposed algorithm with the real base station data in Beijing.The result shows that the map-matching accuracy of our algorithm is higher than that of the prior art.The thesis conduct a congestion analysis based on the result of map matching.The user traffic mode is judged through the extraction of user trajectory data,the shortest path algorithm and trajectory similarity matching algorithm.We judge the possible traffic modes,including walking or riding,automobile and subway.We store users' trajectory data in database,and establish user space location index,so as to realize the crowd flow analysis,spatial query of specified range and space-time compound query of designated time and space.In this thesis,the above functions are integrated into the road traffic analysis system,operators can automatically obtain the user's actual location information,determine the traffic mode and conduct traffic analysis.