A fuzzy-neural scheduling system for achieving on-time delivery in a wafer fabrication factory

This paper constructs a fuzzy-neural scheduling system to improve the performance of job scheduling for a wafer fabrication factory. First, by controlling the standard deviation of cycle time, the fuzzy-neural scheduling system facilitates the rapid assessment of the due date. Subsequently, based on precise cycle time estimation with a fuzzy back propagation network (FBPN), the system attempts to assign a tight due date if more time is allowed. After the jobs related with an order are released into the wafer fabrication factory, a bi-objective dispatching rule is used to shorten cycle time standard deviation, and at the same time, ensure on-time delivery, which is distinct from the previous studies. According to the experimental results, the proposed methodology is better than some existing approaches in both due date assignment and on-time delivery.