“…According to principle, the robust optimization design can be divided into two categories: one category is based on the experimental design of robust optimal design, namely the traditional robust design optimization, such as the main S / N ratio [5], response surface method [6] and, method of loss function [7], the other is modern robust optimal design which is based on the mathematical model of robust optimal design, has mainly the weight method [8], compatibility decision making problems [9] and physical programming [10]. The first class of method is very effective to the test design, but S / N ratio of in the actual engineering is not easy to obtain, and cannot establish a strict mathematical model for robust design; the second kind of method is based on the computer aided engineering (CAD)and nonlinear programming foundation, mathematical modeling is relatively easy, and the system can be reliability into robust design.…”