Publication

Abstract : Electric vehicle (EV) charging infrastructure strategic placement is necessary to facilitate urban fleet electrification, although its effects on planning are complicated due to the convoluted relationships between facility sites and traffic congestion. Conventional planning models often use Euclidean distance measures or fixed network models, which do not model non-linear delays caused by link saturation in high-density metropolitan networks. In order to overcome these constraints, this paper introduces a congestion-conscious framework of planning that joins a bi-level minimization approach with high-fidelity data of road networks. The higher-level problem selects the location of the stations and tries to minimize the overall system-wide weighted travel time, whereas the lower-level problem represents the driver route choice by considering congestion-sensitive performance functions on links. It implements its framework with real-world road topology based on OpenStreetMap of Bengaluru, India that determines a network of more than 470,000 road segments and is topologically valid. The simulation assesses the effectiveness of coverage given limited budgets on station and a lower vehicle range. Findings show that demand-weighted selection uses the focus on high-intensity commercial corridors rather than geometry centrality and has demand-weighted centration of 3.60 km. Moreover, the analysis demonstrates that the chosen structure serves the whole area within a 15 km radius, and it proves that the consideration of the actual road geometry and the amount of traffic catered is essential when planning an actual infrastructure.