基于Kriging模型的涡轮风机转子结构优化及噪声分析OA

To investigate the effects of air flow rate and noise on ducted fan performance,a turbine ro-tor was used to replace the traditional multi-blade centrifugal rotor.Five key structural parameters of the turbine rotor were optimized by combining the surrogate model with genetic algorithm,aiming to further improve the performance of the turbine fan.100 sets of samples were generated by the Latin hypercube experimental design method,and the regression Kriging model was established by taking the maximum airflow of the turbine fan as the objective parameter and combined with computational fluid dynamics(CFD)simulations.Subsequently,a genetic algorithm was used to optimize the regression Kriging model.The results show that the regression Kriging model can accurately describe the relationship be-tween the key design parameters of the rotor and the maximum airflow of the turbine fan.Specifically,the inlet diameter Dj of the turbine rotor and the outlet angle β2 of the blade exert a significant influence on the maximum airflow of the fan,and with the increase of the value,the functional force of the rotor is enhanced.The blade bending degree f exhibits a moderate influence,while the number of blades N and the inlet angle β1 of the rotor have little influence on the maximum airflow of the fan.After the op-timization,the rotor increases the maximum airflow of the fan by 22.4％at the same speed,and re-duces the noise by 4.00 dB at the same airflow conditions.