comparative analysis of metaheuristic approaches for optimal routing of electric vehicles with simultaneous pickup and delivery

Given growing ecological concerns and the push for sustainable transportation, electric vehicles (evs) have emerged as an eco-friendly alternative to conventional vehicles. however, optimizing ev routes, especially for simultaneous pickup and delivery (spd) services, remains a significant challenge requiring advanced solution approaches. this paper presents a comparative analysis of metaheuristic methods for the electric vehicle routing problem with simultaneous pickup and delivery (ev-spd). a comprehensive mathematical model is formulated, incorporating practical constraints such as vehicle capacity and battery energy limits. three metaheuristic algorithms—genetic algorithm (ga), particle swarm optimization (pso), and bat algorithm (ba)—are implemented and compared. pso and ba use a priority-based encoding where a continuous vector is mapped to a customer sequence, partitioned into routes via a greedy capacity-based heuristic. battery constraints are managed using a simple route-splitting repair mechanism. computational results on randomly generated instances show that all three algorithms yield high-quality solutions in reasonable time. the comparative analysis reveals that pso achieves the best balance between solution quality and efficiency, providing the optimal result in 4 out of 6 test cases and consistently yielding the lowest routing costs for larger-scale problems.