A comparison between inverse problem and least square approaches on solving robust design problem by simulation

摘要

Inverse Problem (IP) has been played an essential role in development of modern science and engineering in last several decades. This study performs a comparison between Inverse Problem (IP) and ordinary least squares (OLS) based estimation methods applied on solving robust design (RD) problem by simulation. The results suggest that IP can be a potential alternative to OLS as an estimation tool when empirical models representing relationships between output quality characteristics and input control factors are built. Solution of IP approach can provide a better insight of RD problem, also.