An Integrated Multi-Criteria Decision Making Approach to Location Planning of Electric Vehicle Charging Stations

Hu-Chen Liu, Miying Yang, MengChu Zhou, Guangdong Tian
2018 IEEE transactions on intelligent transportation systems (Print)  
Electric vehicles (EVs) are recognized as one of 1 the most promising technologies worldwide to address the fossil 2 fuel energy resource crisis and environmental pollution. As the 3 initial work of EV charging station (EVCS) construction, site 4 selection plays a vital role in its whole life cycle, which, however, 5 is a complicated multiple criteria decision making (MCDM) 6 problem involving many conflicting criteria. Therefore, this paper 7 aims to propose a novel integrated MCDM approach by
more » ... a grey 8 decision making trial and evaluation laboratory (DEMATEL) 9 and uncertain linguistic multi-objective optimization by ratio 10 analysis plus full multiplicative form (UL-MULTIMOORA) for 11 determining the most suitable EVCS site in terms of multiple 12 interrelated criteria. Specifically, the grey DEMATEL method is 13 used to determine criteria weights and the UL-MULTIMOORA 14 model is employed to evaluate and select the optimal site. 15 Finally, an empirical example in Shanghai, China, is presented to 16 demonstrate the applicability and effectiveness of the proposed 17 approach. The results show that the proposed approach is a 18 useful, practical, and effective way for the optimal location of 19 EVCSs. 20 Index Terms-Electric vehicle, site selection, uncertain linguis-21 tic variables, MULTIMOORA, multiple criteria decision making. 22 On the other hand, there may exist complicated and interre-71 lated relationships between evaluation criteria in a practical 72 EVCS site selection. Decision-making trial and evaluation 73 laboratory (DEMATEL) [18], [19] is an effective method 74 to analyze the inter-relationships among system factors and 75 visualize them by using a cause-effect relationship diagram. 76 Moreover, it is capable of dividing interrelated criteria and 77 dimensions into cause and effect groups [20]. Since its intro-78 duction, the DEMATEL method has been successfully applied 79 in various fields [21]-[27]. Given its strengths, this paper 80 will utilize the DEMATEL to model the dependency among 81 EVCS site selection criteria and further determine their relative 82 weights. 83 With the motivations stated above, this work proposes an 84 integrated MCDM approach based on grey DEMATEL and 85 uncertain linguistic MULTIMOORA (UL-MULTIMOORA) to 86 optimally locate public charging stations for EVs. The main 87 contributions of this study are threefold: First, the theory of 88 uncertain linguistic variables is used to manage the decision 89 makers' uncertain and diverse linguistic assessments. Second, 90 the causal relationships and interaction levels among evalua-91 tion criteria are addressed using the grey DEMATEL method. 92 Third, with the UL-MULTIMOORA model, the proposed 93 approach can get a robust ranking of candidate sites and 94 identify the best one to implement a public EVCS. Finally, 95 an empirical example is presented to demonstrate the potential 96 and advantages of the proposed EVCS site selection frame-97 work. 98 The rest of this paper is structured as follows: We review 99 the EVCS locating literature and indicate research gaps in 100 Section II. The basic definitions and concepts of grey theory 101 and uncertain linguistic variables are recalled in Section III. 102 A hybrid MCDM approach is developed in Section IV for 103 the EVCS site selection. Section V examines the feasibility 104 and effectiveness of the proposed approach by applying it to a 105 practical case. Finally, main conclusions and future directions 106 of this research are presented in Section VI. 107 II. LITERATURE REVIEW 108 Depending on various objectives, a number of MCDM-109 based location models have been proposed in the literature. 110 On the one hand, multi-objective decision making (MODM) 111 techniques have been applied for site selection especially 112 for the deployment of public charging infrastructures. For 113 example, Tu et al. [7] developed a spatial-temporal demand 114 coverage approach for optimizing the placement of electric 115 taxi charging stations considering temporal constraints such as 116 electric taxi range, charging time, and capacity of charging sta-117 tions. He et al. [28] incorporated institutional and spatial con-118 straints, such as local government requirements on charging 119 facility deployment and spatial distribution of potential sites, 120 into facility location models. Shahraki et al. [29] proposed 121 an optimization model based on vehicle travel data to capture 122 public charging demand and applied it to Beijing, China by 123 maximizing the amount of vehicle-miles-traveled being electri-124 fied. Cavadas et al. [30] developed an improved mixed integer 125 828 http://ieeexplore.ieee.org/document/7995120/ 829 Hu-Chen Liu (S'12-M'13) received the M.Sc. 830 degree in industrial engineering from Tongji Uni-831 versity, Shanghai, China, in 2010 and the Ph.D. 832 degree in industrial engineering and management 833 from Tokyo Institute of Technology, Tokyo, Japan, 834 in 2013. He is currently a Professor with the School 835 of Management, Shanghai University. His main 836 research interests include quality and reliability man-837 agement, artificial intelligence, Petri net theory and 838 applications, and healthcare operation management. 839 He has authored over 90 papers in these areas in 840 leading journals, such as IEEE TRANSACTIONS ON FUZZY SYSTEMS, IEEE 841
doi:10.1109/tits.2018.2815680 fatcat:wbcjujcmqbe6pds5isaty5usum