7. CONCLUSIONS
In the present paper, the capacitated location problem in reverse logistics is analyzed. This problem is a MILP with two levels: supply points ( , reprocessing facilities ( ) and ? ∈ ?) ? ∈ ? demand points ( ). The objective is to identify the optimal sites to install reprocessing ? ∈ ? facilities to minimize the variable costs of transportation and management as well as the fixed costs of installing these facilities. To solve the problem, we propose a Benders’ decomposition algorithm and a combination of Benders’ method and tests for fixing variables in a single two-steps algorithm. A set of 60 instances was randomly generated, and the performance of both algorithms is evaluated using this set of problems. It became clear that the performance of the Tests+Benders algorithm is superior to Benders’ method in terms of computation time and effort. The difficulties imposed by the generated dataset are diversified and the efficiency of the proposed algorithm is demonstrated, considering that this method allows all of the testing problems to be solved, including those that could not be solved directly by the CPLEX solver and Benders’ algorithm without reduction tests. In addition to computational aspects, the present study has social relevance: it provides models and methods to help public, private and mixed organizations plan their reverse logistics networks. Any efforts in these areas, whether in the remanufacturing, recycling or reconditioning of products, prevent the generation of waste. Although some companies try to implement the reverse flow of products with the goal of complying with current legislation (Fleischmann et al., 2000), if the business plan is well executed, this segment can become a very lucrative activity for companies (Blackburn, 2004).
