A fuzzy tailored nonlinear fluctuation smoothing rule for job dispatching in a semiconductor manufacturing factory

Job scheduling in a semiconductor manufacturing factory is a difficult task, mainly due to the complexity of the production system and the uncertainty involved in the production activities. Recently, a number of advanced dispatching rules were proposed, which estimate the remaining cycle times of jobs. This predictive nature is conducive to the effectiveness of these rules. However, if the uncertainty in the remaining cycle time can be better considered, then the possibility of incorrect scheduling will be further reduced. To this end, an effective fuzzy aggregation mechanism is established to enhance the performance of the existing fuzzy c-means and back propagation network approach. Subsequently, the two- factor tailored nonlinear fluctuation smoothing rule for mean cycle time (2f-TNFSMCT) is modified in this study, by diversifying the slacks of jobs. The effectiveness of the proposed methodology is illustrated with a simulation study. According to the experimental results, slack diversification was indeed a good idea in improving the scheduling performance of a fluctuation smoothing rule.