Abstract: In order to reduce the aerodynamic noise of a fan, takes a certain axial flow fan as the research object and proposes a noise reduction method based on genetic algorithm, which takes the minimum turbulent kinetic energy on the blade surface as the optimization objective. The noise reduction mechanism and aerodynamic performance of the optimized blade profile are analyzed. Firstly, by parameterizing the blade structure, automatic control of spatial size and shape was achieved. Secondly, with the constant air volume as the constraint objective, three structural parameters are selected, namely the axial bending parameter, the angle between the tangent and chord of the leading and trailing edge points, and combined with sensitivity factor analysis, the response relationship between the turbulent kinetic energy on the blade surface and the blade structural parameters is established. Finally, the NSGA-II genetic algorithm was used to determine the optimal blade structure, and the differences in aerodynamic performance and noise levels before and after optimization were compared to verify the feasibility of the method. The results showed that: (1) Different axial bending shapes have different weights on the impact of air flow and noise; (2) Two optimization schemes were selected: blade #1 can achieve noise reduction by reducing the rotational speed, while ensuring that the air volume is level with the original blade; Compared to the original blade, the air volume of blade #2 is basically the same, and the maximum noise reduction reaches 1.6 dB(A) while ensuring constant speed; (3) Through experimental verification, it was found that the impact of air volume on noise is greater than that of rotational speed. Under the premise of ensuring constant air volume, the effect of noise reduction through blade shape is better than that of reducing rotational speed.

Key words: Axial flow fan, Pneumatic performance, Multi-objective optimization