Comparative Performance Analysis of Relocation Policies for Electric Vehicle Sharing Systems

This paper selects possible relocation scenarios for electric vehicle sharing systems and measures the service ratio through our analysis framework. The experiment exploits the actual trip records collected in a taxi telematics system in Jeju city area as future sharing demand. Its goals include testing the validity of the analysis framework for a sharing system design as well as discovering the behavioral characteristics in basic relocation policies to take into account when designing a new sophisticated relocation scheme. The sharing performance is measured for even, utilization-based, and morning-focused schemes based on the assumption that the relocation is carried out during the non-operation hours. The performance analysis results find out that it is useful to focus on the pick-up requests during the first few hours after the operation starts and that it is necessary to develop a prediction-based proactive relocation scheme to cope with request fluctuation in the airport area.