Site selection for electric vehicle charging stations in car sharing systems: Integration of photovoltaic systems and upgrade decisions

This paper addresses the increasing demand for electric vehicle charging stations (EVCS) in the context of the growing electric vehicle (EV) market and the shift towards clean mobility solutions. We propose a Mixed-Integer Linear Programming (MILP) model to optimize the location of charging stations within an electric vehicle sharing system (EVSS), while also considering necessary upgrades for the stations and the integration of photovoltaic (PV) systems. By focusing on these elements, the model aims to enhance efficiency and sustainability, contributing to the broader goals of reducing environmental impact and promoting alternative fuel vehicles. To address challenges in large-scale optimization, we also propose a heuristic approach using Variable Neighborhood Search (VNS) for scenarios where the MILP model is not computationally tractable. The study was conducted over a seven-year planning horizon and validated the model through a case study in three cities in Saudi Arabia. The performance and efficiency of the developed algorithm are compared with a state-of-the-art commercial solver. The heuristic approach demonstrated improvements of 4.53% over the commercial solver. Both methods were completed within the same total time frame of 24 h, showcasing the heuristic's effectiveness in generating competitive solutions efficiently. Additionally, the study recommends strategic upgrades for PV systems and charging stations, offering insights into the optimal number of each type of charger and number of PV panel to be installed at each location. These findings support data-driven investment decisions in EV infrastructure. The model can assist stakeholders in prioritizing locations and technologies that yield the greatest operational and environmental benefits.