Tri-Level Dynamic Pricing for Charging Network Operator in Coupled Power-Transportation Networks

As a promising way to achieve energy transition, electric vehicles (EVs) are rapidly advancing. Dynamic charging price of charging network operator (CNO) guides the owners to vary travel decisions, which in return affect the operation of power distribution networks (PDNs) and transportation networks (TNs). Therefore, an appropriate dynamic charging price can decrease the operating costs of PDN and TN, while increasing the profit of the CNO. This study proposes a tri-level dynamic pricing model. At the upper level, the CNO sets the charging price to maximize its profit considering the decision of EVs. And at the middle level, the gasoline vehicles (GVs) and EVs make their travel and charging decisions to minimize their travel cost by considering simultaneous charging, routing and departure time (SCRDT) choices, yielding a mixed dynamic user equilibrium (MDUE) pattern. Consequently, the charging loads are determined. At the lower level, the distribution network operator (DNO) calculates the power purchase price by solving a second-order cone ACOPF (SOC-ACOPF) problem with charging loads with temporal and geographical variations. Furthermore, a mixed genetic algorithm nested with the projection gradient method (PGM) and second-order cone programming (SOCP) is designed to solve the proposed model. Numerical examples validate the effectiveness of the proposed model, and analyze the impact of dynamic pricing on the operation of PDN and TN and CNO’s profit.