DISPATCHING RULE SELECTION USING ARTIFICIAL NEURAL NETWORKS FOR DYNAMIC PLANNING SCHEDULING

To schedule a job shop, the first task is to select an appropriate sch eduling algorithm or rule. Because of the complexity of scheduling pro blems, no general algorithm sufficient for solving all scheduling prob lems has yet been developed. Most job-shop scheduling systems offer al ternative algorithms for different situations, and experienced human s chedulers are needed to select the best dispatching rule in these syst ems. This paper proposes a new algorithm for jobshop scheduling proble ms. This algorithm consists of three stages. First, computer simulatio n techniques are used to evaluate the efficiency of heuristic rules in different scheduling situations. Second, the simulation results are u sed to train a neural network in order to capture the knowledge which can be used to select the most efficient heuristic rule for each sched uling situation. Finally, the trained neural network is used as a disp atching rule selector in the realtime scheduling process. Research res ults have shown great potential in using a neural network to replace h uman schedulers in selecting an appropriate approach for real-time sch eduling. This research is part of an ongoing project of developing a r eal-time planning and scheduling system.