Multiobjective Optimization for the Impeller of Centrifugal Fan Based on Response Surface Methodology with Grey Relational Analysis Method

The theories of weighted grey relational analysis combined with RSM are used to optimize the centrifugal fan impeller parameters. In order to evaluate centrifugal fan efficiency (Eff) and total pressure (Tp), two objective parameters Eff and Tp are selected as multiple responses to optimize. The parameters evaluated are inlet blade angle, outlet blade angle, and blade number. Simulations format are based on Box-Behnken design (BBD) method and the values of multiple responses are acquired using computational fluid dynamics (CFD). The values of grey relational grade (GRG) for the multiple responses are acquired based on weighted grey relational analysis. The particularity of weighted grey relational analysis is that the weight factors can be calculated quantificationally from the weightiness of the multiple responses. RSM is used to construct the regression model between GRG and input variables. Based on weighted grey relational analysis and RSM, the optimal centrifugal fan impeller parameters can be obtained, and the acquired results indicate that both Eff and Tp increase after optimization. The results also indicate that the proposed optimization method is a very useful tool for multiobjective optimization of centrifugal fan impeller parameters.