This paper intend to increase the effciency of the prediction of the Response Surface by applying orthogonal and rotatable design in Dural Respones Surface Methodology.

Based on the analysis and comparison of these two methods,the paper presents a double response surface method to combine RSM and Taguchi Methods in order to make use of the advantages of RSM and Taguchi Methods.

This paper analyzes and summarizes the methods for multiresponse optimization,and the theories and applications of the methods based on Mahalanobis distance,multivariate loss function,desirability function,proportion of conformance and dual response surface are addressed.

Current researches on nonparametric dual response surface methodology (NPDRSM) focus on impro- ving fitting performance,but lose sight of improving generalization performance of models.

In this paper, we point out some limitation of dual response surface methodology, and introduce the application of response model methodology in product parameter design improvement with its advantage.

Desulphurization efficiency with spray tower by response surface methodology;

Optimization of antagonism activity yield by Lactobacillus casei sudsp. rahamnosus Ls-81 by response surface methodology;

Optimization of medium composition for expression of recombinant human nerve growth factor(rhNGF) in Escherichia coli cells using response surface methodology;