Heuristic and metaheuristic structure of response surface methodology in process optimization

Response surface methodology is a sequential process that the control variables are optimized and used in experimental design. But according to the characteristics of the method may be stopped at the local optimal region. In this article two innovative methods to improve RSM performance are presented. The proposed methods reduced the possibility of stopping in the local optimal region and improved performance of number of runs. These methods are based on checking the convexity of response surface. For evaluation of the proposed approaches, a simulated process has been analyzed. Numerical analysis shows better performance of proposed approaches than the basic response surface methodology.