Travel Pattern Recognition Based On Mobile User Location Data

Along with the city traffic management needs and the development of traffic planning requirement, extract traffic information based on analysis city residents travel behavior for the division of transportation structure and urban traffic planning was important project in traffic field. Resident travel behavior was data base and evaluation reference for traffic problem research and discuss. Different resident travel pattern had the kinds of influence for urban environment, and traffic jams situation; determine kinds of traffic vehicle proportion in whole urban traffic structure and distribution of traffic facilities. So, how good recognition the resident travel pattern was very important for urban traffic planning and distribution or using of urban traffic space.Traffic information collected based on mobile phone location data was new traffic information gather technology. The main process of it as follow:First, using mobile phone location technology fix phone user position; Second, combination the road matches method and data mining algorithm for extract traffic information from phone user; At last, judgments the residents travel pattern based on existed traffic information. Compare with existed traffic information collection technology, the new one was didn't need super addition cost for set and maintain. In addition, because the new technology also had large quantities of data, was able to collection information of all-weather advantage, it became hotspot research in gather traffic information field.Because of influence elements from communication network and environment, the mobile location data had a lot of communication jamming data. Consulted the existed mobile communication network Optimization algorithm and urban roads traffic management information, combination existed characteristic of mobile location data, the paper propose a new dispose method for mobile location data. After of experimented and results analysis, prove the method had reject a lot of error data may have disturb to location of mobile phone user. According to the achievement from new dispose method of mobile location data, consulted the mobile phone handover theory and the existed characteristic of data from communication network, direct at the shortage of existed road matches method and propose a new road matches method. The new road matches method combination the mobile phone handover location and fizzy logic theory. Experiment and result analysis based on urban roads network data for this method, calculate and extract travel speed data of mobile phone user. Synthesized achievement of research at above, based on characteristic of resident travel behavior and traffic vehicle transfer features, direct at disvantage of existed resident travel pattern method, the paper propose a new recognition method of resident travel pattern. Experiment and result analysis based on mobile location data and road matches method extract travel pattern from mobile phone user, compare with the GPS data of floating car prove the method.At last, summary the paper research work and achievement look into the future and propose further research and resolve spot.