Performance Analysis of a Tour Scheduler Focusing on Time-dependent Gains for Electric Vehicles

After briefly presenting a tour-and-charging scheduler capable of maximizing the timedependent gains for electric vehicles, this paper conducts an extensive analysis of its performance through a prototype implementation. Main concerns are put on the acquired gain, waiting time, and tour length according to the schedule depth and the number of destinations. Basically, the scheduling service provides an interface for the tour spot manager to upload the time-dependent coupons as well as runs the tour scheduler each time a new request arrives from an electric vehicle. According to the experiment made to run on the real-life geographic tour spot distribution of Jeju City, the proposed scheme takes about 4.8 to 4.9 times as much economic gain as the legacy traveling salesman problem solver. The waiting time approaches the permissible bound specified by the explicit constraint, especially when the schedule acquires more coupons. In addition, the tour length is affected by up to 14.4 %. Here, when many coupons are available, the depth 1 vehicle can monopolize gains, but the next vehicle also takes enough economic gains.