GA with hierarchical evaluation: a framework for solving complex machine scheduling problems in manufacturing

Machine scheduling is one of the most important issues in the planning and operations of manufacturing systems, which has been proved very difficult to solve with analytical methods because of its inherently complex constraints and objective. There do exist some successful approaches for standard n/m/J/Cmax job-shop problems, but they hardly consider the actual situations in manufacturing and are therefore not easy to extend for practical use. In this paper, we proposed an intuitive yet efficient framework for solving complex machine scheduling problems. First, definitions of complex scheduling problems are given and then the principles and components of our framework are described in detail. Our framework was applied in an electronics corporation and generated high-quality schedules. What is more, experiments on a number of established job-shop problem instances have shown that this same framework has better performance than most of the other known heuristics even in traditional n/m/J/Cmax problems.