عنوان مقاله [English]
In industrial environments, scheduling systems often operate under dynamic and random circumstances. In these conditions, it is inevitable to encounter some disruptions and breakdowns which are inherently unexpected events. These disruptions bring about the initial schedule to quickly become infeasible and non-optimal and in need of appropriate revisions and rescheduling methods. We consider a flexible flow shop (FFS) system with stochastic or unexpected disruptions such as the arrival of new unpredicted jobs into the process. The occurrence of disruptions and unexpected events in scheduling problems makes
the obtaining of robust and stable solutions more valuable than the finding of optimal solutions that ignore these disruptions. In the literature, for achieving stable solutions, either iteration-based time-consuming simulation methods or surrogate measures (SMs) have been developed; they proactively provide an approximation of the system's real conditions following the occurrence of a disruption due to of the discrepancies of these measures with their true values; however, they may not show the true performance of the system. In this paper, a new reactive approach is considered to achieve a stable scheduling despite unpredicted disruptions, such as unexpected arrivals of new jobs. In this approach, a multi-objective reactive method based on classical and new performance measures is used to control the effects of disruptions that reschedule the initial plans after any unexpected event. An innovative concept called the ``Stability'' is introduced to reduce the effects of the unexpected disruptions. As the FFS problem is NP-hard, considering that stochastic disruptions increase its complexity, the non-dominated Sorting GA-II algorithm or NSGA-II, which is a very famous multi-objective optimization algorithm, is then applied to solve it. To show the performance of the proposed approach, a case study in petrochemical industry is considered. Computational results indicate that this method produces better solutions compared to the classical scheduling approaches used in this company.