Optimization Design of Multi-Blade Centrifugal Fan Based on Variable Weight PSO-BP Prediction Model and Multi-Objective Beluga Optimization Algorithm

Multi-blade centrifugal fans are the main workhorse of automotive air conditioners, and the performance of these fans affects riding comfort. This article proposes a prediction model and a multi-objective optimization algorithm and applies them to the optimization design of a multi-blade centrifugal fan. A prediction model between the design variables and optimization objectives, named wPSO-BP, is proposed, and the model is more effective than the BP prediction model in predicting fan performance. A multi-objective optimization algorithm, named NSGA-III-LBWO, is proposed and applied to the optimization design of the fan along with the wPSO-BP prediction model. The results indicate that the aerodynamics and noise performance of the optimized fan were improved, which provides a reference for the optimized design of these types of fans.