Optimal EV Charging Network Design: When Users Have Choices

The penetration rate of Electric Vehicles is increasing as expected. To meet the charging needs, numerous charging stations are built, which are small because of space and grid limitation in city. Such a high density distribution of charging stations offers EV users more choices when seeking for charging service. In order to offer users better choices with global knowledge, recommendation systems are developed. However, the cost generated from recommendation system is not considered when designing the charging network, which result in an increase of overall system cost. Therefore, this paper mainly focuses on the charging network design problem under generalized recommendation systems to find a better co-working solution. We first formulate the EV charging network design as a cost minimization problem, considering construction cost, incentive cost for users' to transfer in the charging network, and cost for the loss of users leaving the system, which is a nonlinear mixed-integer optimization problem. In order to solve such problem, we propose a branch and bound algorithm. By combining local search heuristic and cutting plane process, the exact optimal solution can be obtained efficiently. To demonstrate the superiority of the scheme proposed by us, simulation are conducted.