A weighted metric method to optimize multi-response robust problems

In a robust parameter design (RPD) problem, the experimenter is interested to determine the values of control factors such that responses will be robust or insensitive to variability of the noise factors. Response surface methodology (RSM) is one of the effective methods that can be employed for this purpose. Since quality of products or processes is usually evaluated through several quality characteristics or responses, more attention should be paid to multi-response parameter design to improve quality of several responses simultaneously. There are many optimization methods in multi-objective decision-making (MODM) area which could be used for this purpose. In this article, some of these optimization techniques are reviewed and a criterion is considered to determine the optimum control setting factors for multi-response RPD problems. A sensitivity analysis is performed to investigate the effect of different scenarios on the solution