Urban Public Transit System Accessibility Evaluation And Optimization Method

With the rapid development of urbanization in China,the urban traffic demand increases rapidly.Besides,large cities have the high urban population density and high development level of land use.Given the priority of public transit services,the urban transportation can realize the sustainable development.From the perspective of accessibility,this study considered the relationship between the urban public transportation system and land use.Then,the urban public transit accessibility was modeled and optimized.This study summarized the researches about accessibility all over the world.Based on the goal,this study defined accessibility as the ability of a person from the given location to reach the activity locations or attain the opportunities through specific traffic systems.This study systematically analyzed the characteristics of accessibility,including spatial attribute,temporal attribute,and socio-economic attribute.Accessibility is suitable for understanding the operation of transport system,predicting the effects of urban traffic planning,comparing the relative accessibilities of different groups,and selecting the household location.Accessibility is mainly dependent on four kinds of factors,including land use factor,transport system factor,time-space factor,and individual factor.In addition,the pros and cons of different accessibility measures were summarized.Furthermore,this study proposed a comprehensive model to evaluate the urban transit accessibility.The main researches and the innovations of this study are as follows:(1)Based on the travel data of urban public transportation,this study extracted the trip data from the metro OD(origin-destination)flow data,the bus transaction data and the taxi GPS(Global Positioning System)data,and computed the travel displacement from the locations(latitude and longitude)of origin and destination.With the actual displacement data of each mode,the maximum likelihood estimation was applied to estimate the parameters of power-law distribution,exponential distribution,lognormal distribution,Gamma distribution and Weibull distribution.Finally,the Akaike information criterion was utilized to select the most fitting distribution.The results demonstrated that metro trip displacements follow the Gamma distribution;bus trip displacements follow the exponential distribution;and taxi trip displacements follow the exponential distribution.Besides,the trip displacement data are selected randomly according to the annual passenger volume of each mode to construct a fusion trip displacement dataset,which follows the power-law distribution in the first part and the exponential distribution in the second part.As fusion trip displacement datasets were different,five fusion trip displacement datasets were constructed to demonstrate the stability of the fitting parameters.(2)This study utilized the taxi GPS data to evaluate the accessibility of the urban area.Firstly,the urban area of Beijing was divided into grids,which were 1km×1km in size.The taxi moving signals and opportunities were allocated to the corresponding grids to calculate the value of opportunity.The taxi GPS data was utilized to attain the travel time between grids,which was used in the cumulative opportunity measure to calculate the accessibility.In the cumulative opportunity measure,three kinds of opportunities were applied,including the number of points of interest,the number of the total taxi drop-offs,and the number of the dynamic taxi drop-offs.In order to reflect the dynamic changing of accessibility in a day,four time periods were selected to calculate the corresponding accessibility,including the morning peak hours,the midday hours,the evening peak hours,and the midnight hours.The results showed that accessibilities in different time periods differed significantly.The midday periods and the midnight periods owned higher accessibility.The accessibilities in the morning peak hours and the evening peak hours were lower.For the three accessibility measures,the grids with high accessibility were located in the core area of the city.In general,the inner part had higher accessibility than the outer part,and the eastern part had higher accessibility than the western part.This study innovatively introduced the number of total taxi drop-offs and the number of dynamic taxi drop-offs as opportunity indicators to measure accessibility,which made it possible to measure accessibility with the taxi GPS data independently.The method can help to evaluate dynamic accessibility more conveniently.(3)Based on the urban rail transit network in Beijing,this study modeled the topology network and proposed a two-stage method to evaluate the accessibility of the metro system.Firstly,the calculative opportunity measure was used to compute "to metro accessibility",which denoted the number of points of interest within the given distance from the given metro station.Secondly,the gravity measure was used to compute "by metro accessibility",which used the travel time between stations as the travel impedance and "to metro accessibility" as the opportunity indicator.Besides,the impacts of bike-metro integration on urban rail transit accessibility were analyzed.The results showed that the stations with high "to metro accessibility" were located in the urban area,and the outer stations were lower.After adopting shared bikes as the feeder mode,"to metro accessibility" increased significantly,especially the stations in the fringe.The stations with high "by metro accessibility" were clustered in the core area of the city.The bike-metro integration can improve "by metro accessibility" significantly,and the stations in the fringe had higher increase rate.(4)This study aimed to model and optimize the travel time and accessibility of feeder bus.In the rectangle service area,four kinds of feeder bus operation strategies were modeled and compared,including fixed route feeder bus with fixed stops,fixed route feeder bus without fixed stops,demand responsive feeder bus with loop line,demand responsive feeder bus with separate route in the two running directions.The analytical modeling was used to compare the travel times and the simulation experiments were used to compare the accessibilities of different operation strategies.The results showed that with the increase of the ratio between length and width,the number of stops for the fixed route feeder bus with fixed stops when the accessibility was maximized would increase.When the travel demand was low,the demand responsive feeder bus with fixed stops had the highest accessibility.When the travel demand was high,the demand responsive bus with separate route in the two running directions performed best in accessibility.As the ratio between length and width increases,the dominant interval of demand responsive bus with separate route in the two running directions will expand.Besides,the impacts of a on feeder bus accessibility were analyzed.This study focused on the modeling of the urban public transit accessibility.Firstly,understanding the travel pattern of public transit riders and modeling their mobility.Secondly,evaluating the accessibility of the three modes,including taxi,urban rail transit and feeder bus.Lastly,from the perspective of accessibility improvement,this study analyzed the impacts of bike-metro integration on the urban rail transit accessibility,and selected the best operation strategy for feeder bus in terms of accessibility.This study will help to promote the sustainable development of urban traffic,which has a profound significance for the long term progress of the city.