Optimal Operation for Cross-Period Scheduling of Electric Vehicles in Traffic-Power Systems

The growing use of electric vehicles (EVs) strengthens the interdependency between the transportation network and the power distribution network, especially in the charging environment of dynamic wireless charging technology. The large-scale EVs’ access to the coupled networks will bring new operational challenges to the traffic-power system, especially during peak hours. To improve the operation of the traffic-power system, this paper adopts a cross-period scheduling mode, which provides users with economic incentives to enable users to participate in scheduling. The operating performance of the traffic-power system is reflected by the total generalized social cost (TGSC), which includes the sum of EV travel time cost, the sum of EV energy consumption cost, management cost, generation cost, power loss cost, and voltage variation cost. Accordingly, to model the minimization problem of TGSC, a bi-level optimization framework is presented. Specifically, the upper level established an optimization problem with TGSC as the objective and road tolls as decision variables. In the lower-level traffic allocation problem (TAP), a multi-period user equilibrium (MPUE) is introduced; i.e., EV users can choose their departure times, thus forming shiftable traffic demands. We verify that the MPUE-TAP formed by the nonlinear complementary problem can be equivalent to a convex optimization problem. In addition, the lower level involves the optimal power flow problem. Due to the complexity of the bi-level programming studied in this paper, we propose an algorithm that combines the surrogate model-based algorithm and adaptive particle swarm optimization to address the optimization problem. Finally, the results of the case studies demonstrate the effectiveness of the proposed cross-period scheduling and the superiority of the proposed algorithm.