Metaheuristics for scheduling on parallel machines under deteriorating and maintenances

In this paper, we consider the parallel machine scheduling problem with the deteriorating effect and maintenances to minimize the makespan. The involvement of condition based maintenance activities not only gives a potential to construct better models and more accurate decision support systems for real-life issues, but it also implies difficulties on search space traversal techniques to guarantee feasibility of results and not omitting optimality. Therefore, we propose well known and reliable metaheuristics that proved their efficiency for various combinatorial optimization problems, i.e., simulated annealing and tabu search. However, our implementations are based on a representation that due to multiple redundant solutions gains its significant advantage to cover all feasible schedules, therefore, the related optimization algorithms guarantee that the optimum is not excluded from the search space. The efficiency of our approach is verified on the basis of computational experiments.