The Research On Information Extraction And Calculation Of Public Transit Service Based On Smart Card Data

The scientific extraction and application of public transit service information is an important measure to improve the operation efficiency and scheduling.It embodies the modern public transit management and service level of a city.At present,the manual survey method is mainly used for the service information data extraction.It is difficult to guarantee the continuity of the manual survey data because of the time-consuming and complicated organization.As the intelligent public transit system application goes deeper,it accelerates the spread of smart card.The use of smart card not only offers beneficial prices for public transit passengers,but also provides a large number of scientific data for the support of public transit service information extraction and calculation.In order to improve the operating efficiency and service level of the public transit system,this paper analyzes the extraction of smart card data(SCD),the ordinary bus arrival time calculation,the travel time calculation of multi-mode public transit and the service reliability evaluation to study the extraction and intelligent application of smart card in detail.For the public transit SCD extraction,the extraction method of smart card information is proposed by comparing the characteristics of manual data,GPS data and SCD.SCD is chosen as the main data source in this research.According to the data composition and the meaning analysis of each data field,the public transit service information,which including the bus arrival time calculation,the travel time calculation and service reliability evaluation extraction steps,are summary analyzed with the required data fields to improve the accuracy of the data extraction.The bus arrival time calculation model established considers the passengers' distribution characteristics and the swiping regulation.Many passengers usually swipe the card much before their alighting.The charging system does not record the boarding time.The directly use of SCD to calculate the bus arrival time will have great errors.Through matching 1 011 sets of SCD and 360 sets of the manual data by the station code,the distribution characteristics of the alighting swiping time and the change of load factor are regressed and clustered.The calculation model of the bus arrival time and the analysis model of the swiping card error value are established and validated for its veracity and feasibility.The error value and the error rate are 0.7% and 3.8% respectively.It proves that the model has high accuracy.As for the public transit travel time calculation,a public transit travel time calculation model is built appropriate for multi-mode travel.The individual travel chain data is matched based on the smart card code.Combined with the arrival time calculation results,the operation characteristics of the metro and the bus rapid transit(BRT)are analyzed and studied to build the travel time calculation model of multi-mode public transit.There are 150 216 sets of data including 5 metro lines and 7 bus lines used for study.The contour line is used to calculate the travel-time graph of the central business district to large residential areas.For the public transit service reliability evaluation,an evaluation system is proposed considering the station coverage ratio,the load factor and the travel time reliability which is based on the calculation results of the travel time.Through the passenger satisfaction manual data result,the station coverage degree,the comfort in the coach and the travel time stability are determined to be three main influence factors.By the data normalization and grade evaluation,there are 157 870 sets of data extracted including 5 metro lines,3 BRT lines and 3 ordinary bus lines.The service reliability evaluation system is built using the central business district(CBD)to the large residential areas for example.The operation status and passengers' travel regularity are analyzed to achieve the intelligent service of the public transit based on SCD.The accurate calculation of the arrival time and the travel time provides the scientific data support for reasonable scheduling.It has the important practical significance to evaluate the multi-mode public transit service reliability scientifically for the management department to optimize the network and promote the passengers' service level.In the future,based on this study,the calculation accuracy of the bus operation information can be further improved with the real-time returned data.The improvement of the public transit information service can not only rise up the quality of life of the residents,but also advance the intelligent public transit system and benefit to the smart city.