Structure Optimization In Axial Flow Fan Impeller Based On The Neural Network And Genetic Algorithm

As the pollution of the environment, especially the increasing air pollution, people are working to develop different devices to improve air quality. Wherein the sprayer has been widely applied in the dedusting and governance haze. Axial flow fan as a core component of sprayer, plays an important role to improve the sprayer performance.Rapid development of CFD numerical simulation technology in the application of axial flow fan structure optimization greatly saved the development cost of axial flow fan.In this paper, by using the NACA four-digit airfoil built the 3-D model of axial flow fan impeller, and determined the main structural parameters of axial flow fan model by theoretical analysis. Using CFD software for simulation analysis of different axial flow fan impeller structure model. When analyzing the impact of the fan duct structure to the axial fan parameters focuses on the maximum camber, the location of maximum camber,the maximum thickness as a fraction of the chord, the chord length and the blade angle.This has a certain guiding significance of the axial flow fan structure design.This paper combined BP neural network and genetic algorithm. Firstly a three-layer BP neural network was established which has five input and one output. And through the use of L-M algorithm realized the fitting of the mapping relationship between axial flow fan impeller structure and range. On this basis, using neural network in genetic algorithm and obtained the optimal structure of the axial flow fan impeller. Eventually the fan performance has been greatly improved.At the same time, this paper compared the results of orthogonal test with the optimization results mentioned above, And analyzed the reason for the difference between their optimization results.