The use of Taguchi, ANOVA, and GRA methods to optimize CFD analyses of ventilation performance in buildings

We present a method that can be used to optimize CFD analyses of ventilation performance in buildings; the was successfully applied other disciplines. In method, three statistical methods are - Taguchi ANOVA, and Grey Relational Analysis (GRA) achieve optimization. then model risk draft mean age air two-dimensional benchmark test case exemplify application method. use five independent variables for modeling: temperature, velocity, room dimensions, each with levels. short, is implemented as follows: At first, using an L25 orthogonal array we decreased total number cases modeled from 3125 (55) 25. Then, 25 ANOVA find best solutions draught or air, while GRA found solution both them considered jointly. Finally, new models describing were run confirm their superiority within selected modeling array. They not compared against all possible because computational time would have been prohibitive. Nevertheless, believe, results credible identifying solutions. This based on experience We, therefore, propose presented when complex problems analyzed related requiring large power.