查看文章

This paper focuses on the optimization of mixed-fleet bus scheduling (MFBS, with vehicles of different sizes) in public transport systems. We develop a novel Mixed-Integer Nonlinear Programming (MINLP) model that addresses the MFBS problem by optimizing vehicle assignment and dispatching programs. The model considers user costs and operator costs, as well as the inconvenience of users from crowding. To enhance the optimization process, we develop two hybrid metaheuristic algorithms, Genetic Algorithm combined with Simulated Annealing (GA-SA) and Grey Wolf Optimizer combined with SA (GWO-SA), with a Taguchi approach to calibrate the metaheuristics’ parameters. The performance of the metaheuristics is extensively evaluated on small, medium, and large-scale samples, considering solution quality and CPU time. Results show that the GWO-SA outperforms the other metaheuristics. Applying our model to a real bus corridor in Santiago, Chile, we find that precise dispatching plans generated by more advanced algorithms (GA-SA and GWO-SA) lead to cost savings and improved performance compared to simpler algorithms (GA and GWO). Utilizing more accurate/advanced algorithms makes a difference in terms of quality of service and optimal fleet size in crowded scenarios, whereas for low and medium demand cases, simpler dispatching algorithms could be used without a drop in accuracy. Overall, our research advances the understanding and optimization capabilities of mixed-fleet bus scheduling, contributing to the efficient and comfortable operation of public transportation systems.