A Real-Time Production Scheduling Method Using Attractor Selection

The real-time production scheduling problem treated in this research is the problem of assigning sequential jobs to plural processing machines with different ability to adjust to such changes of "environment" as job tendencies and production site situations. In this research, a solution is scheduled in real time following environmental changes by switching scheduling strategies, giving priority to a low-cost processing machine or to an empty processing machine using an "assignment parameter". "Attractor selection", which adopts the concept of "attractor (stable state)" chiefly defined in biological and physical fields, is applied for the adjustment of the assignment parameter in changing environments. Scheduling following environmental changes in real time is achieved by transiting the assignment parameter with a transition formula of attractor selection. The proposed method was applied to a real-time production scheduling problem, and schedule optimality and follow-up ability were analyzed when environmental changes occur