Integration of Response Surface Methodology with Genetic Algorithms

Response surface methodology (RSM) is a methodology that combines experimental designs and statistical techniques, for empirical model building and optimisation. By conducting experiments and applying regression analysis, RSM seeks to relate a response to some input variables. This work aims at integrating response surface methodology with genetic algorithms (GAs) to realise a GA-based prototype system for the determination of near optimal values in response surface designs. A framework of the prototype system is presented. The prototype system was validated using three case studies of a bonding process that involve solving the Himmelblau function, optimising the mean pull strength, and maximising both the mean pull strength and the minimum strength simul-taneously. The results were compared with those obtained by the Design Expert, which is a commercial software package. Details of the case studies as well as the comparative studies are presented.