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 simultaneously. 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.