Station Mtching And Bus Arrival Time Deduction Based On IC Card Data

In recent years,China's economy has achieved rapid development,but the problem of urban traffic congestion also follows.The best way to solve this problem is to develop public transport and advocate public transport.With the development of public transport,the traffic data has been accumulated in a large amount,and the research of data mining in the field of public transport has become increasingly hot.At present,there are a lot of researches in the field of public transport at home and abroad,and researchers have achieved enough depth and breadth in the exploration of traffic data.In the field of public transport,especially in the field of public transport,there is a strong dependence on IC card data and GPS data.But in fact,GPS data of public transportation is often missing.Aiming at this problem,this paper studies how to identify the bus stop for IC card data and deduce the method to restore the bus arrival time when GPS data is missing.This paper analyzes the IC card data,and finds that the boarding site of users in a time period is basically fixed.According to this rule,this paper proposes a method to identify the boarding site by the boarding users.Firstly,this paper constructs the characteristic user table based on the historical IC card data that has been used to identify the boarding station,which records the information of the user's swiping behavior.Then,the hierarchical clustering of IC card data is carried out to gather the users of the same station in a shuttle bus into one group,and the voting algorithm is used to score and vote according to the users of this category combined with the characteristic user table to get the most likely boarding station for the users of this category.According to the characteristic user table,the accuracy of the matching results is 83%;this paper continues to analyze the results of the matching errors,summarizes the causes of the matching errors,and finally according to the timing logic combined with the causes of the matching errors,the results are checked again to get more accurate results.Experiments show that the accuracy of the site matching method can reach over 94%.After the identification of the bus stop is completed for the IC card data,this paperproposes a method of bus arrival time restoration based on this IC card data.First of all,this paper analyzes the number of stops with user boarding behavior for each bus,and concludes that it is reasonable to deduce and restore the bus arrival time based on IC card data.Then analyze the bus travel time between stops,and get the characteristics that affect the travel time between stops.At last,the paper deduces the bus arrival time for the stations with users on board by using IC card data.For the stations without users on board,the paper deduces the bus arrival time according to the bus arrival time for the stations with users on board by combining the historical data method,historical data mean method and support vector regression machine(SVR).Through the comparison of experimental data,it can be seen that the IC card data combined with SVR model to deduce the bus arrival time has a higher accuracy.