This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE TRANSACTIONS ON POWER DELIVERY 1 Traf fic-Constrained Multiobjective Planning of Electric-Vehicle Charging Stations Guibin Wang, Zhao Xu, Senior Member, IEEE, Fushuan Wen, and Kit Po Wong, Fellow, IEEE Abstract—Smart-grid development calls for effective solutions, such as electric vehicles (EVs), to meet the energy and envi- ronmental challenges. To facilitate large-scale EV applications, optimal locating and sizing of charging stations in smart grids have become essential. This paper proposes a multiobjective EV charging station planning method which can ensure charging service while reducing power losses and voltage deviations of distribution systems. A battery capacity-constrained EV flow capturing location model is proposed to maximize the EV traffic flow that can be charged given a candidate construction plan of EV charging stations. The data-envelopment analysis method is employed to obtain the final optimal solution. Subsequently, the well-established cross-entropy method is utilized to solve the planning problem. The simulation results have demonstrated the effectiveness of the proposed method based on a case study consisting of a 33-node distribution system and a 25-node traffic network system. Index Terms—Charging station, cross-entropy, data-envelop- ment analysis, distribution systems, electric vehicle (EV), locating and sizing, traffic flow. I. INTRODUCTION I N RECENT years, the climate change has aroused interna- tional awareness about the negative impacts of using fossil fuels. Development of electrical vehicles (EV) industry is re- garded as an important measure to reduce the carbon emissions. With the development of power electronics and battery tech- nology, the EV ownership number gained a rapid growth given governmental supports and incentives like tax reductions. Par- ticularly in China, ambitious plans have been made to take the lead of future EV implementation in the world [1]. At the initial stage, barriers to successful deployment of EVs at large scale exist in various aspects. Among others, lack of Manuscript received November 27, 2012; revised April 11, 2013 and May 26, 2013; accepted June 05, 2013. This work was supported in part by National Basic Research Program (973 Program) (No. 2013CB228202), in part by Hong Kong Polytechnic University Grants (51107114, 51177145), in part by Hong Kong Polytechnic University Grants (G-U962, A-SA73), and in part by the Shenzhen Government Fundamental Research Program (JC201006040906A). Paper no. TPWRD-01285-2012. G. Wang and F. Wen are with the Department of Electrical Engineering, Zhe- jiang University, Hangzhou, Zhejiang 310027, China (e-mail: wgbzju@gmail. com; fushuan.wen@gmail.com). Z. Xu is with the Department of Electrical Engineering, The Hong Kong Poly- technic University, Hong Kong, China (e-mail: eezhaoxu@polyu.edu.hk). K. P. Wong is with the University of Western Australia, Perth WA 6009, Aus- tralia (e-mail: kitpo@ieee.org). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPWRD.2013.2269142 sufficient charging infrastructure is the most critical one. Gen- erally, EV charging infrastructure can be of different types in- cluding mainly battery swapping station and fast charging sta- tion. It is widely believed that fast charging stations have the po- tential to be widely used [2], and consequently more and more attentions have been paid to the optimal planning of them in the past few years [3]–[7]. A dynamic traffic network method is used to build a model with a time constraint to obtain the op- timal location and size of EV charging stations in [3], in which the investment and charging costs are to be minimized. In [4], a two-step method considering environmental factors and service radius is presented to identify the candidate EV charging sta- tions. The resultant multiobjective function considering invest- ment and operation costs, as well as the network losses, is solved by the interior point algorithm. In [5], the feasibility of opti- mally utilizing the potential of the Ontario’s grid for charging plug-in hybrid EVs (PHEVs) is analyzed for off-peak load pe- riods by employing a simplified zonal model of the Ontario’s electric transmission network and a zonal pattern of base-load generation capacities for the years from 2009 to 2025. In [6], an environmentally and economically sustainable in- tegration of PHEVs into a power system is addressed under a robust optimization-based planning methodological framework taking the constraints of both power systems and transport sec- tors into account. In [7], a smart load management approach for coordinating multiple plug-in EVs chargers in distribution sys- tems is proposed, with the objectives of shaving peak demands, improving voltage profiles and minimizing network losses. Research on EV charging station planning is still in the early stage. There are still many influential factors that should be incorporated in planning charging stations. These include the EV owner’s driving behavior, the topologies of distribution system and the traffic network, and the economics and security issues of power systems. These factors have only been partially addressed in most existing publications. In this work, the power distribution and traffic network topologies and the EV owner’s driving behavior are taken into account and a new multiobjec- tive charging station planning method is formulated. Because of its robustness and fast convergence, the cross-entropy method is utilized to solve the multiobjective planning model and ob- tain the Pareto planning solutions. A novel data-envelopment method is then used to make the final planning decision among the Pareto solutions to determine the optimal charging station location and its size simultaneously. The effectiveness of the proposed method has been demonstrated through a case study based on a test system consisting of a 33-node distribution system and a 25-node traffic network system. 0885-8977/$31.00 © 2013 British Crown Copyright