Tabu-based GIS for solving the vehicle routing problem

• An integrated sytem that merges optimzation rules and a geographical information system.
• A real case study applied on the area of Tunisia shows the performance of the platform.
• A tabu search approach applied on benchmark instances highlights the out performance of its results.
• A comparative study with state of the art tabu search metaheuristics is driven.

Besides being a hard combinatorial problem, the VRP is also a spatial problem. Hence, effective decision making in this field strongly requires the integration of GIS and optimization systems (GIS-O). This article integrates GIS and optimization tools for solving the vehicle routing problem with loading and distance requirements (DCVRP). A general outline of the multi-step integration is pointed out showing the interaction of the GIS and the spatial optimization according to the loose coupling strategy. The computational performance of the TS-VRP algorithm for the DCVRP turned out to be quite efficient on both computation time and solution quality. The Tunisian case study well illustrates the incentive behind using such a spatial decision support system that allows the management of the problem from the data acquisition to the visualization of possible simulation scenarios in a more realistic way.