Iterated Local Search Algorithm for Flexible Job Shop Scheduling

Abstract

The flexible job shop scheduling problem (FJSSP) is an extension of the classical job shop scheduling problem (JSSP). The classical JSSP determines the job sequence for each machine. In contrast, FJSSP decides not only the job sequence but also the machine assignment for jobs; thus, it is a combinatorial optimization problem of a larger scale. To efficiently solve an FJSSP, which is non-deterministic polynomial-time hard, we require a heuristic method. In previous studies, the FJSSP has been solved by neighborhood algorithms, which employ various metaheuristic methods. Some approaches constrain the neighborhood operation from jobs on a critical path and simultaneously change the machine assignment and job sequence. However, a method that can independently change machine assignment and job sequence may improve the efficiency of FJSSP because the solution of JSSP already exists. In this study, we investigate the effect of continuing to change job sequence and machine assignment, and job sequence and machine assignment change are iterated using a local search method. In numerical experiments, the effects of the frequency of job sequence change and machine assignment change on the performance of the solution are investigated. Finally, we find the best machine assignment and job sequence change method for FJSSP.