Comparação entre estratégias para programação integrada de operações envolvendo decisões de programação de operações da produção e roteirização de veículos

Abstract

The main motivation of this paper is to study a solution refinement technique for scheduling and routing of integrated production and distribution problems. It is possible to affirm that solving these operations in an integrated form is important to guarantee the search for the best solution for the system. The objective of the research is to compare the results obtained from three types of resolution: a mathematical model solved in an uncopled form, an integrated model and an integrated model that uses an initial solution of the uncoupled problem. The main characteristics of the studied problems were: unique machine and one vehicle for distribution. As the objective function, was prioritized the minimization of the total flow time of the orders in the system, normally related by the literature with the total time of fulfillment of an order and with inventory levels. To achieve this goal, mixed integer linear programming was used and two different time limits were set, 120 and 1.800 seconds. One of the most important conclusions was that, with a relatively short time limit of execution, 30 minutes, the use of an uncoupled initial solution in an integrated model, in groups of highest number of jobs, provides an average improvement of approximately 11.15% in results compared to the integrated model and 18.24% of improvement when compared to the uncoupled form.