A Solution Proposal for the Capacitated P-Median Problem with Tabu Search

Abstract

This work presents the Capacitated P-Median problem, which seeks to solve the optimal location of p distributors. In understanding the computational complexity of this problem, partitioning principles are applied in the process of associating a distribution center with customers to be serviced. These are treated as clusters containing each distribution center and its customers. The optimization of the implicit cost function within the partitioning (distance minimization) is performed using Tabu Search with a diversified search. A series of computational experiences is performed aided by the OR-Library test instances, where in most cases an optimal solution is reached in reasonable computing time. As Tabu Search does not obtain all expected optima, OR-Library test instances are solved with the help of Lingo 16.0 and the obtained results are compared with those resulting from Tabu Search.