EV charging schedule for load leveling by non-linearly-distributed start time

Large-scale deployment of electric vehicles (EVs) adds load to the current power system. Without control of the temporal dispersion of EV charging, a new rapid peak load may occur and cause severe problems in the power system. Therefore, we propose two algorithms for load leveling by decentralized autonomous control. The first algorithm uses an off-peak rate period, and the second changes the charging start time according to the charging duration. We evaluated the second algorithm in two cases: a linear case, where the start time varies linearly with the charging duration, and a quadratic case, where the start time varies non-linearly, that is, quadratically with the charging duration. In the linear case, a gradually sloped peak occurs in the morning. On the other hand, in the quadratic case, the charging hours are dispersed appropriately, and the daily load curve is almost flat in the night.