Applying Grey Relational Analysis and Fuzzy Neural Network to Improve The Yield of Thin-Film Sputtering Process in Color Filter Manufacturing

The paper proposes an approach to improve the yield of Chrome (Cr) thin-film sputtering process of the black matrix (BM) in color filter manufacturing. The grey relational analysis and fuzzy neural network (FNN) are used to find the robust parameters of the process with multiple quality characteristics. The parameter optimization process was taken by performing the Taguchi method. Two quality characteristics and four experimental factors with three levels were selected and an L9 orthogonal array was chosen to conduct experiments. Base on the results of grey relational analysis with entropy weight, the optimal process parameters were obtained. The confirmation test showed that manufacturing with the optimal process parameters reduced the opportunities of repair and rework of Cr thin-film and raises the yield. This research also utilized the fuzzy neural network to learn and extract the rules in the BM production. The 3D response surface generated from the fuzzy rules can provide the engineers with process window to support their decision making on process parameters.