This study focused on the uncertain flexible flow shop scheduling problem with limited buffers when preventive maintenance is applied at fixed intervals. This issue has not been addressed in spite of widespread applications, due to complexity arising in solving such a stochastic decision making problem. To this aim, a novel optimization model is presented along with two types of solving methods using metaheuristic algorithms with and without a computer simulation model. The proposed hybrid method, named HSIM-META, integrates the computer simulation model along with the three most common metaheuristic algorithms, i.e., genetic algorithm (GA), simulated annealing (SA) algorithm, and particle swarm optimization (PSO), which offer better solution quality according to the literature. For this purpose, the simulation outputs are applied as an initial population for the tuned metaheuristic parameters to look for the next improved solution by investigating different approaches. Different numerical examples are discussed to examine the performance of the proposed method. The computational results of the proposed method, including hybrid simulation with GA (HSIM-GA), SA (HSIM-SA), and PSO (HSIM-PSO), are compared with the just applying GA, SA, and PSO. The results reveal that the suggested method acts more efficiently in terms of accuracy and speed in solving the problem.
An Improved Genetic-Simulated Annealing Algorithm Based on a Hormone Modulation Mechanism for a Flexible Flow-Shop Scheduling Problem