Modelling Battery Electric Vehicle Sharing’s Use Behavior And Estimating Its Potential Demand Shift

In order to deal with issues such as traffic congestion,environmental pollution,and energy consumption,transport management departments apply many countermeasures.Among them,battery electric vehicle(BEV)sharing is taken as one of the sustainable transport modes and starts to enter the markets of main cities in China.However,due to the primary launch in the Chinese market,there is still lack of corresponding research in both methodology and application.To be specific,it is still not clear about users’ preferences in both adoption and mode choice in particular trips.Besides,the potential demand for BEV sharing as well as the impact of BEV sharing on the urban transport system are not studied thoroughly.Therefore,in order to provide references for BEV sharing system planning and management,this study aims to study the use behavior of BEV sharing facing the urban transport system.From the angles of the mid-term and short-term decision-making processes,this study models the adoption intention of BEV sharing and its mode choice behavior.Further,based on the real trip data,the potential demand shift is estimated to evaluate the impact of BEV sharing on the urban transport system.In detail,the main works and findings are summarized below.(1)Investigations on the attitude and choice preference related to BEV sharing.In order to provide data support for BEV sharing use behavior modelling,based on designing the stated preference survey(SP),the information about respondents’ personal and household attributes,car ownership status,travel patterns,attitudes towards BEV sharing and scenario-based mode choice preference are collected.By statistical analysis,the most possible potential trip scenario of using BEV sharing is leisure trip within 10-20 km.Respondents who are not locally registered and don’t own private vehicles are likely to use BEV sharing.(2)Modelling BEV sharing adoption considering attitudes.From the perspective of the mid-term decision,the adoption intention is studied.In the modelling,in order to promote the model’s performance and explanatory ability,the attitude factors are considered in addition to traditional variables such as level of service(LOS),personal attributes,trip scenarios.Based on the statistics of attitudinal indicators,31 indicators are used to construct the factor structure for latent attitudinal variables using factor analysis.In order to consider the endogeneity between attitudinal variables and utility,as well as relieve the bias brought by exogenous variables,a Hybrid choice model is incorporated and simultaneously estimated: a multiple-indicator multiple-cause model is incorporated to capture the relationship between observed variables and latent variables,and a multinomial logit model is used to comprehensively investigate the effects of the key factors on carsharing choice.Results indicate that attitudes such as environmental consciousness,social benefits,satisfaction with transport system,and reliability significantly affect the adoption of battery electric vehicle sharing.Results also indicate that the restrictions on vehicle use owing to license plates improve the adoption of battery electric vehicle sharing.(3)Modelling the mode choice behavior of BEV sharing.In order to provide insights into the behavior mechanism of travelers from the perspective of short-term decisions and analyze the mode preference for BEV sharing comparing with other existing modes,it is essential to study the use behavior of BEV sharing in trip level.To clarify the influence of special factors such as access/egress distance and driving range,the shared vehicle choice problem is also considered into the framework of mode choice model.To alleviate the effects caused by Independence from Irrelevant Alternatives(IIA),a Nested Logit(NL)model is developed to jointly analyze the mode choice and route-vehicle choice under various trip scenarios.In the model,LOS variables,personal attributes,and trip scenarios are considered.The error components are introduced to solve the problems caused by serial-correlated data.The impacts of the level-of-service variables including access distance and remaining range are investigated.Besides,in order to have a deep insight into choice preference,the perceived utility is tested by designing nonlinear forms of functions and perceived thresholds.Results show that taxi is the main competitor with BEV sharing,especially for long-distance trips.Access distance to a shared BEV has a more significant impact on travelers’ mode choice.It is also found that when the trip distance increases,travelers tend to become more tolerant of the access distance and require a higher remaining range of shared BEVs.(4)Analysis on potential demand shift to BEV sharing.In order to handle the characteristics of potential demand shifts and the impacts on the urban transport system,based on the estimated NL model,real trip data is incorporated to estimate the potential demand shift for battery electric vehicle sharing.The temporal and spatial distribution of potential demand shifts,the impact of battery electric vehicle sharing on the mode split,and the impact of pricing strategies are analyzed.In addition,according to the trade-offs between access/egress distance and cost,an incentive policy facing user-based relocation is designed in order to reduce the tasks in relocation,and further tested in a case study.The results show that an optimistic mode split of battery electric vehicle sharing is 4.23% when the average distance between travelers and stations is 0.5 km.The main source of potential demand shift is public transport(7.66%).This implies that large-scale development of BEV sharing may aggravate traffic congestion.