Robust parameter design for nonlinear signal–response systems using kriging models

Abstract

ABSTRACT Most of the literature on robust parameter design concerns simple response systems; however, many engineering systems can be described as signal–response systems, in which the relationship between the signal factor and response is linear or nonlinear. In this article, a robust parameter design method is proposed to optimize nonlinear signal–response systems. To make the signal–response relationship insensitive to noise factors, the variation model and response model are developed using kriging models. In the variation model, the mean and standard deviation of the variation in response over the range of the signal factor are fitted and then synthesized as the robustness criterion, which is minimized to optimize the settings of control factors. The response model is used to obtain the fitting model of the signal–response relationship at the optimal setting of control factors. The applicability and efficiency of the proposed method are demonstrated using an engineering case.