A Maximum Cluster Algorithm for Checking the Feasibility of Dial-A-Ride Instances

The dial-a-ride problem (DARP) involves the dispatching of a fleet of vehicles to transport customers requesting service and is one of the most challenging tasks of combinatorial optimization. We study the DARP as a constraint satisfaction problem, where the goal is to find a feasible solution with respect to the time, capacity, and precedence constraints, or to prove infeasibility. The main contribution of our work is a new robust method for this problem formulation. The algorithm is based on a dynamic subroutine that finds for any set of customers a maximum cluster, that is, a maximal set of customers that can be served by a single vehicle. The performance of the algorithm is analyzed and evaluated by means of computational experiments, justifying the efficiency of the solution method.