为了提升低转速工况下压气机的气动性能，采用人工神经网络与遗传算法相结合的优化方法对某单级离心压气机离心叶轮的弯特性进行优化计算。利用NUMECA软件对该离心压气机进行了不同转速的数值模拟，得到压气机不同工况下的气动性能。通过设置不同控制参数和曲线形式对离心叶轮叶片进行参数化拟合，以8个改变叶片弯特性的参数为自由参数进行了叶型优化设计，最终得到了优化后的叶轮叶片。结果表明:优化后在低转速的设计工况下离心压气机压比增加了4.69%，稳定裕度拓宽了17.41%。

In order to improve the aerodynamic performance of the compressor at low speed,an optimization method based on artificial neural network and genetic algorithm is used to optimize and calculate the bending characteristics of a certain singlestage centrifugal impeller.NUMECA software is used to numerically simulate the centrifugal compressor at different speeds,and the aerodynamic performance of the compressor under different working conditions is obtained.The centrifugal impeller blades are parameterized and fitted by setting different control parameters and curve forms.The blade profile optimization design is carried out with 8 parameters that change the bending characteristics of the blades as free parameters,and the optimized impeller blades are obtained.The results show that after the optimization,the pressure ratio of the centrifugal compressor is increased by 4.69%,and the stability margin is widened by 17.41% under the design conditions of low speed.

TK474.8

国家科技重大专项(2017-I-0007-0008)