Desirability Functions in Multiresponse Optimization

Abstract. Desirability functions (DF)s play an increasing role for solving the optimization of process or product quality problems having various quality characteristics to obtain a good compromise between these characteristics. There are many alternative formulations to these functions and solution strategies suggested for handling their weaknesses and improving their strength. Although the DFs of Derringer and Suich are the most popular ones in multiple-response optimization literature, there is a limited number of solution strategies to their optimization which need to be updated with new research results obtained in the area of nonlinear optimization. In this study, we elaborate the piecewise differentiable structure of the DFs of Derringer and Suich.