Research And Design Of Traffic Pattern Recognition Based On The Smart Phone

With the popularity of mobile Internet, various mobile Internet applications contain a huge number of business opportunities. For example, the automatic identification of transport pattern obviously has a broad application value in the field of traffic planning and Location Based Service. At the present stage, smart phone has more and more powerful functions, so communication function is no longer the only one. Smart phone not only has powerfully lain in communication function, but also a strong perception, computing and storage capacity. However, in the current study about the traffic pattern recognition based on smart phone, single source of data is difficult to make the traffic pattern recognition algorithm have a good adaptability. In addition, the pattern recognition algorithm also has different adaption ability in different scenarios, which becomes a big obstacle to improving the traffic pattern recognition accuracy. Hence, a method based on smart phone and random forest algorithm, to identify the traffic pattern recognition, is proposed in this paper. This paper mainly includes the following parts:Firstly, a variety of data sources are used to solve the problem of the single source of data in the traffic pattern recognition, and then the influence of using the new data source on the recognition accuracy is studied.Secondly, the influence of different traffic pattern recognition algorithms on recognition accuracy is researched, especially the comparison of the similarities and differences between the performance of the random forest algorithm and the support vector machine algorithm.Thirdly, the design of traffic pattern recognition system based on smart phone mainly includes overall design, detailed design, etc. In terms of overall design, the data acquisition, feature extraction and model construction based on random forest are detailed explored in this system. In terms of the detailed design, the design of system interaction, storage, data acquisition, feature extraction, pattern discrimination and master control are completed.Fourthly, the traffic pattern recognition based on smart phone mainly includes the operating environment of system hardware and software, the realization of system function module, the construction of test environment and the recognition result of traffic pattern, etc.In sum, the traffic pattern recognition system based on the smart phone is researched and designed in this paper. The recognition accuracy for several kinds of traffic patterns can reach 90% after testing. The testing turns out that the system has a very good value of application.