This master’s thesis addresses the Two-Stage Capacitated Facility Location Problem (TSCFLP), which involves choosing the location of factories and warehouses with capacity and demand constraints. This is a highly relevant problem in the context of smart cities, which use advanced technologies to optimize service provision. The objective of the problem is to reduce the costs of opening facilities and reducing product flow. To solve this problem, this work proposes a new hybrid methodology, called GSA-CS-QVND-LB, which combines the use of Generalized Simulated Annealing (GSA) and Clustering Search (CS) metaheuristics to identify and cluster high-quality solutions. After clustering these solutions, local search strategies guided by a Reinforcement Learning algorithm are applied, followed by Local Branching. The results obtained through computational tests presented high-quality solutions in competitive computational time compared to the most recent data reported in the literature.
.png)
.png)
