Robust Global Sensitivity Analysis for Robust Design under Parameter Uncertainty

Abstract

Abstract: Based on the theory and method of robust design, the robust global sensitivity analysis of products or systems under parameter uncertainty is discussed. A basic idea of the author is to define the robust sensitivity that is the importance measure of the design variables for product functional response function distribution. The Taylor series of moments of the functional response function is carried out, and the approximate analytical formulas of robust global sensitivity are obtained by using the importance measure model based on variance. Finally, a numerical example is given to illustrate the operation principle of this method, and an engineering example is given to verify the correctness of this method.