SOLVING THE OPEN SHOP SCHEDULING PROBLEM VIA A HYBRID GENETIC-VARIABLE NEIGHBORHOOD SEARCH ALGORITHM

Abstract

Abstract In this article, a hybrid metaheuristic method for solving the open shop scheduling problem (OSSP) is proposed. The optimization criterion is the minimization of makespan and the solution method consists of four components: a randomized initial population generation, a heuristic solution included in the initial population acquired by a Nawaz-Enscore-Ham (NEH)-based heuristic for the flow shop scheduling problem, and two interconnected metaheuristic algorithms: a variable neighborhood search and a genetic algorithm. To our knowledge, this is the first hybrid application of genetic algorithm (GA) and variable neighborhood search (VNS) for the open shop scheduling problem. Computational experiments on benchmark data sets demonstrate that the proposed hybrid metaheuristic reaches a high quality solution in short computational times. Moreover, 12 new hard, large-scale open shop benchmark instances are proposed that simulate realistic industrial cases. Keywords: MetaheuristicsOpen shopProduction managementScheduling This work was supported by the General Secretariat for Research and Technology under contract GSRT-05-ΠAB-71. The authors would also like to thank the anonymous referees for their constructive comments and contribution to the completion of this work. Notes LB: The lower bound of a given instance; Min = min [(min sum of processing times within jobs) ∪ (min sum of processing times within machines)]; WL: The workload of a given instance; Best: The best makespan acquired by PGAVNS; Average: The average makespan by PGAVNS after 20 runs; Time: The minimum computational time required to obtain the best found solution. Where operations 1 to 9 belong to job 1, operations 10 to 18 belong to job 2 and so on so forth.