Optimal placement of electric vehicle charging stations

The rapid adoption of electric vehicles (EVs) necessitates the development of efficient and accessible charging infrastructure. One of the primary challenges in this domain is the optimal placement of charging stations, which directly impacts user accessibility, operational cost, and overall service efficiency. This study formulates the EV charging station location problem as a Mixed Integer Linear Programming (MILP) model, integrating both installation costs and distance-based service costs under demand and capacity constraints. Given the NP-hard nature of the problem, exact optimization methods quickly become computationally impractical as problem complexity increases. To address this limitation, a single-objective Genetic Algorithm (GA) and a multi-objective Non-dominated Sorting Genetic Algorithm II (NSGA-II) are proposed and implemented as scalable heuristic alternatives. The developed algorithms are tested on both real-world and synthetically generated datasets, allowing for a comprehensive evaluation of solution quality, computational efficiency, and algorithmic robustness. Comparative analysis with exact solutions obtained from the CPLEX solver shows that GA provides near-optimal results with significantly reduced computation time. Moreover, NSGA-II effectively captures the trade-off structure between installation and service costs by generating a diverse Pareto front, enhancing decision-making flexibility in multi-criteria planning contexts.This research contributes to the field of EV infrastructure planning by offering a robust and adaptable solution methodology that can be applied across diverse urban environments. Future work may focus on integrating dynamic demand profiles, renewable energy considerations, and interactive decision-support systems to further extend the model's applicability