Optimal Placement of Electric Vehicle Stations Using High-Granularity Human Flow Data

Suboptimal placement of charging infrastructure is a major barrier to the transition to sustainable transportation, even with the growing popularity of electric vehicles (EVs). The research addresses this challenge by proposing a novel hybrid genetic algorithm (GA) to solve the NP-hard Multiple-Choice Multidimensional Knapsack Problem (MMKP) for computationally derived optimal charging station placement and configurations in Sapporo, Japan. The methodology leverages high-granularity human flow data to identify charging demand and a Traveling Salesperson Problem (TSP)-based encoding to prioritize potential station locations. A greedy heuristic then decodes this prioritization, selecting charger configurations that maximize service capacity within a defined budget. The results reveal that as the budget increases, the network evolves through distinct phases of concentrated deployment, expansion, and saturation, with a nonlinear increase in covered demand, indicating diminishing returns on investment. The findings demonstrate the efficacy of the proposed model in providing a strategic roadmap for urban planners and policymakers to make cost-effective decisions that maximize charging demand coverage and accelerate EV adoption.