A note on multi-response robust parameter optimization based on RSM

This paper discusses multi-response robust parameter optimization based on response surface models using desirability function method. When specifications of the output responses are given, the design region may be divided into several areas or spaces. Research shows that solution with highest desirability may not be robust. To achieve robust solution we should consider where the solution lying in the feasible operating regions and analyze the sensitivity of the solution when the values of controllable variables change. A two-factor with three output responses example is given and analyzed in this paper. And a straightforward approach on how to find solutions based on the trade-off between robustness and optimality is proposed with applications of this example.