A Decision Support System Based on Hybrid Metaheuristic for Solving the Constrained Capacitated Vehicle Routing Problem: The Tunisian Case

Abstract

Various metaheuristic approaches have emerged in recent years to solve the capacitated vehicle routing problem (CVRP), a well-known $\mathcal {NP}-$ hard problem in routing. In CVRP, the objective is to design the route set at a lower cost for a homogenous fleet of vehicles, starting from and going back to the depot, to meet the needs and expectations of all the customers. In this paper, we propose an ILS-VND approach which is a hybrid of Iterated Local Search (ILS) and Variable Neighborhood Descent (VND) approaches. Although both ILS and VND approaches, independently provide good solutions, we found that the hybrid approach gives better solutions than either approach independently. We demonstrate the effectiveness of our approach through experiments carried out on widely used benchmark instances. Numerical experiments show that the proposed method outperforms other local searches and metaheuristics. We also, propose a Decision Support System (DSS) that integrates a Geographical Information System (GIS) to solve the problem under scrutiny. In order to demonstrate the performance of the proposed DSS in terms of solution quality, we apply it for a real case on the city of Jendouba in the north west of Tunisia. The results are then highlighted in a cartographic format using Google Maps.