Diseño de un Procedimiento Basado en Algoritmos Genéticos para dar Solución al Problema de Scheduling Admitiendo Rechazos en Escenarios Flexible Flow Shop

Abstract

This research addresses the orders acceptance and Scheduling problem, considering a Flexible Flow Shop environment with non-related machines. The objective is to maximize total revenue, which is composed by the revenue for the accepted orders, the penalties for those which are rejected and the costs of the idle time of the resources. To solve it, in first place the mixed integer program is formulated, which formally represents the environment and solve problems with up to 8 orders and 5 stages. Given its complexity, a heuristic approach is developed which evaluates the convenience of the arrived orders, providing an initial solution to the problem. The proposed procedure is designed with this approach and a genetic algorithm, which aims to take advantages from the recombination of solutions and gives, accordingly to the computational tests, solutions for instances up to 30 orders and 10 stages in reasonable time.