Um algoritmo híbrido busca tabu/ILS para o problema de sequenciamento da produção em ambiente open shop

Abstract

The open shop scheduling problem is a production environment in which the jobs are processed in m machines where all the machines can process any of the n jobs and each machine presents a specific processing time for a given job. Since the OSSP is NP-complete, a tabu search/iterated local search (ILS) metaheuristic is proposed. We carried out computational experiments on 140 literature problem instances in order to test the performance of the proposed algorithm. The well-known priority rules LPT, SPT, LAPT and LTRPOM were evaluated as the initial solution for the proposed meta-heuristic. The objective function adopted was minimizing the makespan, and we used the relative deviation as performance criteria. Computational results point to a very well performance of the proposed meta-heuristic for the tested instances.