An iterated local search for the multi-commodity multi-trip vehicle routing problem with time windows

The Multi-Commodity Multi-Trip Vehicle Routing Problem with Time Windows calls for the determination of a routing planning to serve a set of customers that require products belonging to incompatible commodities. Two commodities are incompatible if they cannot be transported together into the same vehicle. Vehicles are allowed to perform several trips during the working day. The objective is to minimize the number of used vehicles.We propose an Iterated Local Search that outperforms the previous algorithm designed for the problem. Moreover, we conduct an analysis on the benefit that can be obtained introducing the multi-trip aspect at the fleet dimensioning level. Results on classical VRPTW instances show that, in some cases, the fleet can be halved.