A hybrid iterated local search metaheuristic for the flexible job shop scheduling problem

In the flexible job shop scheduling problem (FJSP) we have a set of jobs and a set of machines. A job is characterized by a set of operations that must be processed in a predetermined order. Each operation can be processed in a specific set of machines and each of these machines can process at most one operation at a time, respecting the restriction that before starting a new operation, the current one must be finished. Scheduling is an assignment of operations at time intervals on machines. The classic objective of the FJSP is to find a schedule that minimizes the completion time of the jobs, called makespan. Considering that the FJSP is an NP-hard problem, solution methods based on metaheuristics become a good alternative, since they aim to explore the space of solutions in an intelligent way, obtaining high-quality but not necessarily optimal solutions at a reduced computational cost. Thus, to solve the FJSP, this article describes a hybrid iterated local search (HILS) algorithm, which uses the simulated annealing (SA) metaheuristic as local search. Computational experiments with a standard set of instances of the problem indicated that the proposed HILS implementation is robust and competitive when compared with the best algorithms of the literature.