Generating feasible schedules for a pick-up and delivery problem
Résumé
In this research, we study a transportation problem that involves vehicle routing and driver scheduling for a bus station. The problem requires drivers to provide pick-up and delivery services to customers. Its solution involves planning itineraries for buses and establishing working schedule for drivers,subject to vehicle capacity limitation and time constraints. The objective is to efficiently schedule the fleet of vehicles for customer demand so as to reduce costs. This paper presents a complete constraint model and a solution method for solving the problem. For vehicle routing, a permutation constraint is used to impose a total order for visiting all customer locations regardless of different vehicle routes. This provides a global planning over all routes and plays an important role in the solution method. For driver scheduling, set partitioning constraints are used for assigning drivers and vehicles to requests. Based on this constraint model, efficient reactive planning and optimization algorithms are constructed to generate feasible schedules for the problem.