作　　者： (叶涛）; (陈飞）;

机构地区： 武汉理工大学,湖北武汉430070

出　　处： 《流体机械》 2017年第8期24-28,58,共6页

摘　　要： 以某型涡轮增压压气机为研究对象,采用逆向工程技术中的三维扫描的方法反求压气机叶轮,建立叶轮的几何模型,在此基础上,对叶轮几何型线进行参数化拟合,进而以效率和压比为优化目标,利用人工神经网络和遗传算法对叶轮进行气动优化设计。结果表明:优化后,叶轮的气动性能得到很大提高,在优化点叶轮的效率比原模型提高了2.01%,压比比原模型提高了0.12,综合稳定裕度也提高了。 A certain type of turbocharger compressor was as the research object,using the method of reverse engineeringtechnology of3d scanning to construct geometric model of compressor impeller.On this basis,the profile line of the impeller wasfitted parametric.And with the aim to optimize efficiency and pressure ratio,artificial neural network(ANN)and genetic algorithmwere used to make aerodynamic optimization design of the impeller.The results show that the aerodynamic performance of impelleris improved greatly after optimizing,and the efficiency of optimized impeller increased by2.01%and pressure ratio increased by0.12compared to original one at the optimized point,and the integrated stability margin is also improved.

关 键 词： 叶轮 逆向工程 人工神经网络 遗传算法 气动优化