Analysis On The Subway With Bus Transfer Time Based On IC Card Data

In recent years, IC card system is widely applied in many domestic cities, the proportion of passengers who use IC card in many large and medium-sized cities has reached 60%-80%. The development of computer technology caused cities to establish the bus IC card information database and management system. The subway has been developed sustainably in China, and the transfer between subway and bus also increased gradually, so perfect bus passenger information can be obtained by using IC card to collect a lot of information to identify and analyze passengers'transfer behavior. In view of this, on the basis of the national key basic research development plan "coordination mechanism and coupling theory of bus leading multimode transport network", this article used a large number of passengers travel information recorded by IC card, and used the data mining technology to process the collected data, and analyze the relationship between information, then recognize transfer behavior between subway and bus to realize the analysis and optimization of transfer time based on the city of Nanjing bus IC card data.First of all, the article introduced the composition of Nanjing transportation IC card system, the structure of transportation IC card data, types and formats of basic data. Then we selected the Access as the data pretreatment platform to process formats of data, and delete redundant and error data, as the basis of data analysis later.Then, using the software of Access and Excel as platform, we identified the transfer behavior between subway and bus by using the large number of information collected by IC card. To improve the accuracy of the identification, we improved the method of unified threshold identification which is common at present, and used the combination method of initial identification and depth of the recognition, to determine transfer behavior. In the meantime, we used different transfer determination threshold for the transfer between subway and different bus routes around subway, and put forward different behavior identification method based on different threshold value.Based above, logarithm normal distribution fitting is adopted to the transfer time and the P-P figure of SPSS software is used to conduct a distribution test. Then use probability integration algorithm to construct total transfer time model and use Weibull distribution to fit the transfer time on foot. On the purpose of making the total transfer time which passengers consume during research period least, with the constraint condition of load factor, bus departure interval and so on, construct a transfer time optimization model, and then use genetic algorithm to solve it, optimizing the transfer between the two transfer modes.In the end, based on IC card data in Nanjing, analyze the characteristics of the transfer between subway and bus, and make a case study with Zhujianglu subway station. By analysis on the IC cards data of Zhujianglu station, the transfer amount has an obvious morning peak between Zhujianglu station and No.6 bus, due to this, set feeder bus routes optimizing the transfer time on the purpose of effectively connecting Zhujianglu station.