Analysis Of Urban Public Transportation For Designing Of Traffic Warning System Based On Data Mining

Major cities in China advocate public transportation priority policy since the Twelfth Five-year Plan. To increase the public transportation rate, the traffic administrative department of the city should improve quality of service. Convenience, speed, punctuality and comfort are the most important influence factors of traffic mode choosing behaviors. Reasonable transportation planning can effectively solve the problem of imbalance between supply and demand, which could optimize traffic service of public transit. Forecasting of passenger flow for traffic network, lines or stations, could be well applied to urban transportation network planning and traffic infrastructure constructions. Meanwhile, shortterm passenger flow forecasting will provide data for the traffic organization. Traffic management departments make contingency plan in advance for sudden large passenger flow caused by critical incidents, holidays, weather and so on.Traditional passenger flow prediction based on data from special passenger flow investigation or household travel survey can’t meet the actual demand, since the small sample size and the long data acquisition cycle. In this paper, forecasting and forewarning of urban public transportation data were studied. Firstly, we obtained large scale of passenger flow information from the fusion of data recorded by bus Intelligent Card(IC) with GPS data recording the bus trajectory. Secondly, we proposed a hybrid-factor prediction model. The factors affection include: weather, holiday and critical events and so on. The RMSE of our model have decreased an average of 13%. Our empirical study shows that the factors such as pressure, temperature and holiday have significant impact on passenger flow. Finally, the public transportation passenger flow forewarning system is constructed supported by big data platform. Our system can forecast the future passenger flow in the urban transportation network, and push notifications for large passenger flow of lines and stations. This would effectively improve the emergency ability of urban public transport system, thus upgrade the public transit service level.