A tabu search heuristic for the vehicle routing problem with backhauls and time windows

This article describes a tabu search heuristic for the vehicle routing problem with backhauls and time windows. In this problem, the set of customers is partitioned into two subsets: linehaul customers where a given quantity of goods is delivered from a central depot, and backhaul customers where a given quantity of goods is collected and transported to the depot. Mixed routes with both linehaul and backhaul customers are constructed. A variant of this problem is examined where linehauls must precede backhauls on each route. Furthermore, the time of beginning of service at each customer location must occur within a particular time interval. Computational results are reported on a standard set of test problems, and comparisons are made with other heuristic approaches and with optimal solutions.