A Fuzzy-neural Dbd Approach for Job Scheduling in a Wafer Fabrication Factory

Job scheduling is an important but difficult task to wafer fabrication factories. To further improve the performance of job scheduling in a wafer fabrication factory, a fuzzy-neural dynamic-bottleneck-detection (DBD) approach is proposed in this study. The fuzzy-neural DBD approach is modified from the traditional DBD approach after incorporating some major changes. First, taking into account the uncertainty of job classification, fuzzy partition is applied to divide jobs into different categories. Second, the fuzzy c-means and fuzzy back propagation network (FCM-FBPN) approach is applied to estimate the remaining cycle time of a job. Third, we replace the heuristics in the traditional DBD approach, with more advanced and flexible dispatching rules, such as the shortest cycle time until next bottleneck (SCNB) rule and the four-factor bi-criteria nonlinear fluctuation smoothing (4f-biNFS) rule. A real wafer fabrication factory is also simulated as a testing environment for the adoption of several methods. According to the experimental results, the fuzzy-neural DBD approach was better than six existing approaches and their variants in reducing the average cycle time and cycle time standard deviation at the same time.