The demand-adaptive system(DAS)has been recognized as a promising transit mode for demand with high fluctuations.In this paper,we optimize the routes and request selection for a DAS with multiple service routes.Curren...The demand-adaptive system(DAS)has been recognized as a promising transit mode for demand with high fluctuations.In this paper,we optimize the routes and request selection for a DAS with multiple service routes.Currently,most studies on DAS focus on optimizing single-route systems,where each area is exclusively served by one route and heuristic pre-assignations of requests are made.In contrast,our study addresses a more generalized routing and request selection problem for a DAS with multiple service routes.This problem jointly assigns requests to the service routes and determines the resulting routes while considering the pickup and delivery locations and the reserved boarding time for each request.A mixed-integer linear programming(MILP)model is developed to minimize the sum of bus travel time cost,passenger in-vehicle and waiting time costs,and request rejection penalties.A tailored adaptive large neighborhood search algorithm(ALNS)solves this optimization model efficiently.The numerical experiments show that,under the same optimality conditions,the proposed algorithm outperforms the exact algorithm implemented by GORUBI in terms of solution quality and computation time.The ALNS algorithm also reports cost reductions of up to 50%in comparison with prevailing benchmark metaheuristics.Moreover,the multi-route DAS in this paper has a lower rejection rate and objective value than the single-route systems examined in previous studies.展开更多
基金supported by the National Natural Science Foundation of China(Grant Nos.72371130 and 72001108).
文摘The demand-adaptive system(DAS)has been recognized as a promising transit mode for demand with high fluctuations.In this paper,we optimize the routes and request selection for a DAS with multiple service routes.Currently,most studies on DAS focus on optimizing single-route systems,where each area is exclusively served by one route and heuristic pre-assignations of requests are made.In contrast,our study addresses a more generalized routing and request selection problem for a DAS with multiple service routes.This problem jointly assigns requests to the service routes and determines the resulting routes while considering the pickup and delivery locations and the reserved boarding time for each request.A mixed-integer linear programming(MILP)model is developed to minimize the sum of bus travel time cost,passenger in-vehicle and waiting time costs,and request rejection penalties.A tailored adaptive large neighborhood search algorithm(ALNS)solves this optimization model efficiently.The numerical experiments show that,under the same optimality conditions,the proposed algorithm outperforms the exact algorithm implemented by GORUBI in terms of solution quality and computation time.The ALNS algorithm also reports cost reductions of up to 50%in comparison with prevailing benchmark metaheuristics.Moreover,the multi-route DAS in this paper has a lower rejection rate and objective value than the single-route systems examined in previous studies.
