Design And Implementation Of The Travel Way Recognition System Based On The Smart Phone

The bus public transport system for JiangMen City has some problems to be solved.The amounts of passengers for each bus lines are extremely uneven,besides the peak effect caused by the work-time and exwork-time also impact the bus public system;especially in the Central Business District,this phenomenon is being amplified.This is why planning the bus lines and optimizing bus public transport system becomes a challenge to the JiangMen Bus Transportation Group.The JiangMen government has required the Bus Transportation Group to take the advantage of the big data,to research people travel modes of transportation,to calculate the route and to research the people transporting pattern.Base on the research,the transportation department of government can re-plan the bus lines to fit people actual requirement.In the meanwhile the government can advise people to choose the most convenient way to their destination base on this data.For achieving this purpose,the JiangMen Bus Transportation Group invites the research team from Beijing Institute of Technology to research and develop the travel way recognition system based on the smart phone.As I am a technician of the Bus Transportation Group to process this project,below are the main researches and developments.1.Introducing various data sources,solving the problem that low accuracy cause by the single data sources.In addition,study the improvement about adding new data sources in the system,will see how many percentage go up on the accuracy.2.Studying the recognition accuracy improvement due to different algorithm,especially compare the Random Forest Algorithm(RF)with the Support Vector Machine(SVM).3.To design the travel way recognition system based on the smart phone,including the concept for system designing,the whole system designing and the system details designing.About the concept for the system,we particularly study the data collecting,the characteristic analyzing and building the Random Forest model.For the system details designing,we finish the interface designing,data storing,data collecting module designing,characteristic analyzing and main control module designing.4.To realize and test the travel way recognition system based on the smart phone;include building the software and hardware.To design the function module and to design the test to verify the system if it is functional are also included in our study.This paper will describe how to develop the traffic pattern recognition system based on the smart phone.This system has been tested it can recognize people travel by walking,running,biking,driving or using public transportation system such as bus and subway.The accuracy of the system is up to 90%,it is already achieve the JiangMen Bus Transportation Group requirement.