A Pool-Based Hybrid Genetic Algorithm with Local Search for Electric Vehicle Charging Scheduling

Abstract

The electric vehicle charging scheduling problem aims to minimize total delay while satisfying capacity constraints and improving the efficient utilization of charging infrastructure. In this study, a NEH-like heuristic is proposed to generate high-quality initial solutions, and a pool-based hybrid genetic algorithm is developed for solution optimization. The proposed approach combines genetic operators, local search, and diversity preservation mechanisms to improve solution quality and reduce premature convergence. The effectiveness of the proposed method was evaluated through multiple experiments on problems of different sizes and compared with FCFS, random solution generation, genetic algorithm, GRASP, and memetic algorithm approaches. Experimental results show that the proposed method consistently achieves the lowest average objective values among the compared methods and becomes increasingly effective as the problem size grows. These findings de monstrate that the proposed approac is a promising solution method for electric vehicle charging scheduling problems.