Optimization models for placement of an energy-aware electric vehicle charging infrastructure

This paper addresses the problem of optimally placing charging stations in urban areas. Two optimization criteria are used: maximizing the number of reachable households and minimizing overall e-transportation energy cost. The decision making models used for both cases are mixed integer programming with linear and nonlinear energy-aware constraints. A multi-objective optimization model that handles both criteria (number of reachable households and transportation energy) simultaneously is also presented. A number of simulation results are provided for two different cities in order to illustrate the proposed methods. Among other insights, these results show that the multi-objective optimization provides improved placement results.     

A review of spatial localization methodologies for the electric vehicle charging infrastructure

With view to the high share of the transport sector in total energy consumption, e-mobility should play an important role within the transition of the energy systems. Policymakers in several countries consider electric vehicles (EV) as an alternative to fossil-fueled vehicles. In order to allow for the development of EV, the charging infrastructure has to be set up at locations with high charging potential, identified by means of various criteria such as demand density or trip length. Many methodologies for locating charging stations (CS) have been developed in the last few years. First, this paper presents a broad overview of publications in the domain of CS localization. A classification scheme is proposed regarding modeling theory and empirical application; further on, models are analyzed, distinguishing between users, route or destination centricity of the approaches and outcomes. In a second step, studies in the field of explicit spatial location planning are reviewed in more detail, that is, in terms of their target criteria and the specialization of underlying analytical processes. One divergence of these approaches lies in the varying level of spatial planning, which could be crucial depending on the planning requirements. It is striking that almost all CS locating concepts are proposed for urban areas. Other constraints, such as the lack of extensive empirical EV traffic data for a better understanding of the driving behavior, are identified. This paper provides an overview of the CS models, a classification approach especially considering the problem’s spatial dimension, and derives perspectives for further research.