In this study, a systematic approach to achieve a globally optimal Chemical
Mechanical Polishing ( CMP) process is carried out. In this new approach,
the orthogonal array technique adopted from the Taguchi method is used to r
ealize an efficiently experimental design. The RBFNF neural fuzzy network i
s then applied to model the complex CMP process. The signal-to noise ratio
(S/N) analysis (ANOVA) technique used in the conventional Taguchi method is
also implemented to obtain the local optimum process parameters. The globa
lly of optimal parameters are successively acquired in terms of the trained
RBFNF network. In order to increase CMP throughput, a two-stage optimal st
rategy is also proposed. Experimental results demonstrate that the two-stag
e strategy performs better than the original approach even though the total
processing time is reduced by 1/6.