Applying GAs to complex problems: the case of scheduling multi-stage intermittent manufacturing systems

Although genetic algorithms (GAs) have been successfully applied to a wide range of conventional scheduling problems, their application to more complex scheduling environments has been minimal. For a GA to succeed, it is usually necessary to evaluate a large number of individual solutions: which in the contest of scheduling means generating and evaluating a large number of individual schedules. For problems of this kind, creating an individual complete schedule requires considerable effort and significant execution time. Hence, for GAs to be effective in such cases without resorting to inordinate computing power they have to be designed to produce high-quality solutions, while examining a relatively small number of individual schedules (solutions). One complex industrial scheduling problem is that of scheduling multi-stage, intermittent manufacturing systems with intermediate storage.