Bus Boarding And Alighting Stations Identification Based On Multi-source Data

In recent years,the automatic fare collection system and automatic vehicle location system,has been widely used.Subsequently,a large number of bus IC card data and bus GPS data are generated,including a large number of passenger travel information waiting to be dug up.Compared with the traditional manual survey method,using multi-source data fusion method to collect passenger travel information can save a lot of manpower and material resources,and can extract more comprehensive travel information,providing scientific basis for urban bus operation and management.Based on this,through the integration of bus IC card data,bus GPS data and the corresponding bus line and station information,a set of boarding and alighting stations of bus IC card passengers identification method is proposed,which mainly includes three parts: the boarding station identification based on GPS positioning,the transfer behavior identification based on the bus transfer line and waiting rule,and the alighting station judgment based on the urban bus trip chains,previous and commuting records.As for the boarding station identification,firstly,the bus arrival data is extracted through GPS data,and the vehicle and line number are used to connect with the bus IC card data.Then,the time matching method of ‘the time difference between the travel time and the previous arrival time and the next arrival time’ is proposed and compared with the time matching method of ‘the arrival time and the departure time’ and of ‘the previous arrival time and the next arrival time’ so as to complete identification of boarding station.Finally,an example from Ningbo bus system is given to verify the method.The example shows that the matching rate of this method is 89.80%,which is higher than the 71.43% of the time matching method of ‘the previous arrival time and the next arrival time’.It further demonstrates the way of time difference,which can solve the problem of partial IC card swipe time offset from GPS time.As for the transfer behavior identification,firstly,the waiting time and the departure of vehicles during the waiting period are estimated by using the lines of passengers’ transfer and the waiting rule.Then,the transfer behavior identification condition is proposed,that is,compare the waiting time and the number of departing vehicles during the waiting period.It is compared with the hypothetical conditions of ‘the on-go trip’ to complete the transfer behavior identification.Finally,an example from Ningbo bus system is given to verify the method.The example shows that the matching rate of this method is 96.97%,which is higher than the 28.79%of the conditions using ‘the on-go trip’ hypothesis.It further demonstrates that the transfer behavior identification condition proposed in this paper has a higher accuracy rate,especially in starting station transfers.As for the alighting station identification,firstly,make a preliminary identification of the trip type.Then,identify the alighting station of successive trips and of the last trip of the day.After that,the assumption of the same travel mode,of the reverse travel mode and of the commute travel mode are supplemented.Based on the assumptions,the alighting station of the single travel record,of the commute record,and of the records that cannot be recognized by the previous method are identified.Finally,an example from Ningbo bus system is given to verify the method.The example shows that the matching rate of the method based on trip chains and history records is 83.33%,and that of the method based on commute records is 83.78%,which can meet the needs of actual bus system planning.Finally,take the identification results of bus passengers’ boarding station and alighting station in Ningbo city during the week of December 4-8,2017 as an example.Study the analysis method of bus passengers’ travel characteristics from the space-time perspective of travel time,travel consumption,amount of card swiped at the station and source or sink station,as an application extension of the above content.The study method in this paper is of extensibility and can also be applied to bus transit system in other cities.