A new approach to solve the flexible job shop problem based on a hybrid particle swarm optimization and Random-Restart Hill Climbing

Abstract Industries are looking for ways to expand their competitive advantages, a way is optimizing their production, and in this context, they found solutions in activities of production scheduling. The production job-shop scheduling can be a complex problem of combination. The Flexible Job-shop Scheduling Problem (FJSP) is an extension of the job-shop problem and has been widely reported in the literature. Thus, new optimization algorithms continues to be developed and evaluated, in special, artificial intelligence algorithms of the swarm type presented favorable results. In the FJSP context, this research presents the resolution of the FJSP multi-objective, using a hierarchical approach that divides the problem into two sub-problems, being the Particle Swarm Optimization (PSO), responsible for resolving the routing sub-problem, and Random Restart Hill Climbing (RRHC) for the resolution of scheduling sub-problem. The implementation of the proposed hybrid algorithm has new strategies in the population initialization, displacement of particles, stochastic allocation of operations, and management of scenarios partially and totally flexible. Experimental results using technical benchmarks problems are conducted, and proved the effectiveness of the hybridization, and the advantage of PSO + RRHC algorithm compared to others local search algorithms in the resolution of the scheduling problem.