| Recently, the application of robust design in industry is growing more and more important, especially, the combination of response surface method (RSM) and robust design is of interest currently. The main contributions of this paper are discribed as follow: (1) The mathmatical models and method of statistical analysis have been set up for robust design including uncontrollable noise factor. (2) The models of response surface for one noise factor and multinoise factors have been set up , and not only the statistical analysis method of models but also distributions and efficiency of some statistics have been presented for robust design of dynamic character. (3) For example of a thermal expansion instrument抯 robust design, the dynamic response surface models have been set up, and it has been proved that the RSM is an improvement to Taugchi抯method in the eye of above example. (4) From theory and application , the comparison between Taugchi抯 method and the RSM has been made, and the result has shown that the RSM has more advantages than the S/N ratio 3 method and is an effective approach improved to Dro.Taugchi抯 S/N ratio method. |