Key points are not available for this paper at this time.
The Vehicle Routing Problem with Time Window (VRPTW) is of crucial importance in modern societies, where the aim is to optimize resource utilization, reduce costs and address constraints such as time and vehicle capacity. Traditional genetic algorithms often face premature convergence and slow speed in solving such problems. This paper proposes an Improved Genetic Ant Colony Optimization (IGA-ACO) algorithm to efficiently solve VRPTW. The algorithm combines the strengths of a genetic algorithm with the Generalized Variable Neighborhood Search (GVNS) and Ant Colony Optimization (ACO), aiming to minimize the total cost and optimize balance. The Solomon insertion heuristic is employed to initialize the population and enhance local search capabilities, while the two-population structure improves global search performance by exchanging the optimal solutions between the two populations, helping to avoid local optima. Experiments on the Solomon benchmark dataset show that the IGA-ACO algorithm matches the Best Known Solution (BKS) in Class C instances, reduces vehicle usage by 24.45% in Class R, with a travel distance deviation of 9.19%, and slightly reduces vehicle usage by 0.19% in Class RC, with a travel distance deviation of 7.05%. These results demonstrate the algorithm’s effectiveness in optimizing vehicle routing, particularly under complex constraints, and its competitive advantage over other methods.
Chen et al. (Fri,) studied this question.