Analysis Of Ride-sharing Potential Based On Bus Data

With the rapid development of China's social economy and urbanization,people's demand for transportation is also increasing rapidly,which is not only reflected in the requirement of time efficiency,but also in the continuous improvement of travel comfort requirements.At present,with the rising of the number of motor vehicles,traffic congestion becomes more and more serious,and the cost of traffic energy consumption and traffic time is on the rise.Therefore,traffic problems have become social problems closely related to each urban resident's travel.How to effectively alleviate the urban traffic pressure and improve the efficiency and the comfort of the travel for people under the premise of making full use of the available social vehicle resources,which has become the traffic problem drawing worldwide attention.In this context,the “ride-sharing” travel service,such as Uber,Didi Chuxing,provides people with a new travel experience of more convenience and comfort.However,although the available ride-sharing platforms have been mature in matching services between "nearby vehicles" and “passengers with active requests",for those with "no-active requests" having potential of this needs,the platforms are seriously lack of the ride-sharing information and its relative analysis especially for the vehicles(or their drivers).In view of this problem,taking into account that the bus system has a large user base,bus travel is still the main way of urban residents travel.Therefore,based on the data of bus GPS and bus card data,this paper analyzes the sharing potential of urban residents,and provides valuable information on the potential shared travel demand for the vehicles,so as to improve the travel efficiency and comfort of urban residents,achieving the goal of alleviate urban traffic pressure.The main research and achievements include:(1)Bus station identification.In the field of research on the use of bus data for individual / group travel behaviors,the effective identification of bus stops is a basic work to study the problems of bus travel.In this paper,bus GPS data is used to propose a method of bus station identification based on bus driving mode.Based on the running status of the bus,this method divides the bus behaviors into the terminal standby mode,the normal driving mode,the passengers getting-on/off in the station mode.And this method extracted bus stops of up/down direction through the driving mode judgment,the track direction segmentation and DBSCAN density clustering.This paper uses the bus GPS data of Shenzhen City to verify the station identification,and the experimental results show that this method has a satisfactory accuracy of the bus station identification.(2)The deduction of stations that passengers getting on / off.Stations passengers get on /off directly affects the starting and stopping locations of the sharing travel,and deduction of getting on/off stops accurately is the basis of analyzing the potential of“ride-sharing”.This paper proposed getting on/off stations deduction method based on data matching.Based on bus GPS data,passengers' card data and bus stops identification,stations that passengers getting on can be deduced through bus matching,time matching and stops matching.Afterwards,the strategy deduces which station do the passenger getting off based on the principle that the stop passengers get off is near stop they get on in the next time.This paper experimented with bus card data and GPS data of Shenzhen City.(3)The potential analysis of ride-sharing.This paper studied the potential of bus passengers whose home and work place are close to each other,which is based on bus data,bus stops identification and deduction passengers getting on/off.In the context of travel time analysis and a passengers' travel model a passenger's home and work place can be divided.In view of whether passenger getting on the bus or not in ride-sharing,a passenger's ride-sharing potential model is constructed.The paper used bus data of Shenzhen city,analyzing the potential of bus passengers whose home and work place are close to each other based on model proposed above,verified the validation of the model,which provided valuable analysis for ride-sharing service.This research improved the efficiency of ride-sharing and alleviate the pressure of urban transportation.