Abstract
Output projection is a critical task for a semiconductor fabrication factory. There are two steps in output projection: predicting the output time for every job in the factory, and projecting the outputs into each period. For enhancing the effectiveness of output projection, a fuzzy‐neural approach is proposed in this study. Firstly, a fuzzy back propagation network (FBPN) is applied to generate fuzzy‐valued output time forecasts. Then, the fuzzy output projection function is proposed to project the outputs into each period. To evaluate the effectiveness of the proposed methodology and to make comparison with some traditional approaches, production simulation is also applied in this study to generate test data.
Notes
Corresponding author. (Tel: 886–4–24517250 ext. 3645; Email: [email protected])