Optimización de costos en la cadena de suministro en empresas de ensamblaje: comparación de algoritmos evolutivos en el caso de estudio de ensamblaje de televisores

Abstract

Optimization of the different resources of the supply chains (CS) of the companies allows having efficient and effective management of them. Generally, this leads to a dilemma because of the different optimization objectives, which in many cases are in conflict. And due to the nature of assembly companies, optimization problems are complex, and metaheuristic techniques are necessary for their solution. The purpose of this work was to develop and solve a CS model for the assembly industry that seeks to optimize two objectives: profit maximization (considering the cost associated with provisioning, production, storage and distribution), and maximizing the level of customer service. The problem was modelled considering multiple products and periods, for two security inventory scenarios, a maximum and a minimum. For this purpose, three algorithms were compared: NSGA-II, MOPSO and MOMA. The multi-objective optimization tool PlatEMO was used to test these algorithms. As main results, the genetic algorithm NSGA-II presented the best results with a 97.87% service level, associated with an increase in profit of $ 391,556.66 concerning the worst result given by MOPSO in the best optimization scenario, with a level Minimum security inventory.