Real-Time Charging Navigation of Electric Vehicles: A Non-cooperative Game Approach Jun Tan and Lingfeng Wang Department of Electrical Engineering and Computer Science University of Wisconsin-Milwaukee, Milwaukee, Wisconsin 53211, USA Emails: juntan@uwm.edu, wang289@uwm.edu Abstract—This paper proposes an integrated electric vehicle (EV) charging navigation framework which takes into consideration the impacts from both the power system and transportation system. The proposed framework links the power system with transportation system through the charging navigation of EVs. It benefits the two systems by attracting EVs to charge at off- peak hours and saving the time of EV owners with real-time navigation. Based on the formulated framework, a non- cooperative game approach is proposed in this study to model the competition between electric vehicle charging stations (EVCSs). The simulation results show that the proposed integrated charging navigation approach is effective in improving both the reliability and economic profits of the power system. Index Terms—Charging station, electric vehicle, charging navigation, game theory, traffic flow. I. INTRODUCTION With the increasing penetration of electric vehicles (EVs) and the development of charging infrastructure, the electric vehicle charging stations (EVCS’s) are becoming a vital recharging source for EVs. Home charging at a house garage may be more convenient for the EV owners. For people living in urban areas with high population density, the accessibility of personal garages is limited and public charging stations are needed to recharge their EVs. Moreover, EVCSs can offer lower charging prices for EVs compared with home charging, as power can be purchased at a lower rate from the wholesale power market [1]. Also, EVCSs are a much needed recharging infrastructure for long-distance travelers who may run out of their batteries before returning home. These merits make the EVCS a promising charging infrastructure. However, as the penetration level of EVs grows, the intermittent charging loads may place additional stress on the power system by overloading the distribution transformers and transmission lines. Various studies have been conducted for investigating EVs’ impacts on the power system [1]-[8]. Most of them assumed that EVs are charged at home, which are focused on controlling the charging process of EVs in order to shave the peak load or improve the power quality [4]-[8]. For the charging station, the charging duration is much shorter than the home charging. With fast charging, EVs can be charged to full state of charge (SOC) within half an hour [1]-[3], while the home charging needs 6 to 8 hours. Thus, mitigating the negative impact of EVs on power systems through controlling the charging duration and charging rate of EVs is not applicable for the scenario of charging stations. One promising solution is to attract EVs to charge at appropriate times so as to optimize the charging load of EVCSs. In this case, the impact from the transportation system is not negligible for the management and coordination of multiple EVCSs. Reference [3] proposed a rapid charging navigation system for EVs based on the power system coupled with the traffic data accounting for the impact due to traffic flow. The power market is also a factor that should be considered in managing EVCSs. Multiple EVCSs in the same area may belong to different owners, so competitions between different EVCSs can be caused. This type of competition has been modeled by game-theoretic approaches in [1], [2]. However, little work has been done to formulate both the traffic flow and the competition of EVCSs into an integrated problem. This paper proposes an integrated charging navigation framework, which is made up of the power system, transportation system, navigation system, EVCSs and EVs. Based on this framework, a non-cooperative game approach is proposed to model the competition between EVCSs and manage them in a decentralized fashion. To solve the formulated problem, a particle swarm optimization (PSO) learning based best response algorithm is developed, which is able to improve the economic benefits and reduce the peak load of the power grid at the same time. This study has made contributions in several major aspects by: (1) proposing an integrated charging navigation framework to link the power system with transportation system; (2) proposing a traffic flow simulation method for EVs considering the real-world usage data of EVs; (3) proposing a novel business model for EVCSs based on game theory; (4) developing an artificial intelligence learning based best response algorithm for the formulated game problem. II. SYSTEM MODELING A. System Architecture To study the impacts from both the power system and transportation system in the EV charging process, we proposed an integrated EV charging navigation framework as shown in Fig. 1. The integrated EV charging navigation 978-1-4673-8040-9/15/$31.00 ©2015 IEEE