A data‐driven statistical approach for extending electric vehicle charging infrastructure. (19th February 2018)

• Record Type: Journal Article
• Authors: Pevec, Dario; Babic, Jurica; Kayser, Martin A.; Carvalho, Arthur; Ghiassi‐Farrokhfal, Yashar; Podobnik, Vedran
• Other Names: Nižetić Sandro guestEditor.
• Is Part Of: International journal of energy research. Volume 42:Number 9(2018)
• Page Start: 3102
• Page End: 3120
• Publication Date: 2018-02-19
• Subjects: charging infrastructure -- data science -- electric vehicles -- green transportation -- energy informatics; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Power resources -- Research -- Periodicals; 621.042
• DOI: 10.1002/er.3978
• Languages: English
• ISSNs: 0363-907X

Abstract

Summary: Current trends suggest that there is a substantial increase in the overall usage of electric vehicles (EVs). This, in turn, is causing drastic changes in the transportation industry and, more broadly, in business, policy making, and society. One concrete challenge brought by the increase in the number of EVs is a higher demand for charging stations. This paper presents a methodology to address the challenge of EV charging station deployment. The proposed methodology combines multiple sources of heterogeneous real‐world data for the sake of deriving insights that can be of a great value to decision makers in the field, such as EV charging infrastructure providers and/or local governments. Our starting point is the business data, ie, data describing charging infrastructure, historical data about charging transactions, and information about competitors in the market. Another type of data used are geographical data, such as places of interest located around chargers (eg, hospitals, restaurants, and shops) and driving distances between available chargers. The merged data from different sources are used to predict charging station utilization when EV charging infrastructure and/or contextual data change, eg, when another charging station or a place of interest is created. On the basis of such predictions, we suggest where to deploy new charging stations. We foresee that the proposed methodology can be used by EV charging infrastructure providers and/or local governments as a decision support tool that prescribes an optimal area to place a new charging station while keeping a desired level of utilization of the charging stations. We showcase the proposed methodology with an illustrative example involving the Dutch EV charging infrastructure through the period from 2013 to 2016. Specifically, we prescribe the optimal location for new ELaadNL charging stations based on different objectives such as maximizing the overall charging network utilization and/or increasing the number of chargers in scarcely populated areas.