Un approccio Matheuristico ad un problema reale di Vehicle Routing

The domain of logistics involves managing the flow of goods to meet the customer needs. The Vehicle Routing Problem (VRP) is a well-known problem in the field of Operation Research, often used to model and efficiently solve logistics problems. Real life VRPs present various combination of constraints to model different scenarios, making them more challenging to manage than classic VRPs described in the literature. This paper proposes a matheuristic based algorithm to solve large-scale instances of a real life VRP enriched with popular constraints such as Time Windows or the use of a heterogeneous fleet of vehicles. The key step of the proposed algorithm is the optimization of a Set Covering Problem, making it highly adaptable to tackle the complexities of real life VRPs with different constraints. The algorithm was tested with benchmark instances up to 1000 customers and compared to the more competitive ALSN algorithm developed by Optit S.r.l.