采用基于神经网络及遗传算法的叶轮机械叶片三维优化设计方法开发高载荷透平动叶片.pdfVIP

Received date ： 2003-04-23 ； Revision received date ： 2003-08-05 Foundation item ： Project G1999022307 supported by 973 ArticIe URL ： http ： / / www. hkxb. net . cn / cja / 2003 / 04 / 0198 / The Development of Highly Loaded Turbine Rotating Blades by Using 3D Optimization Design Method of Turbomachinery Blades Based on Artificial Neural Network ! Genetic Algorithm ZHOU Fan-zhen 1， FENG Guo-tai 1， JIANG Hong-de 2 （ 1 . Academy of Energy Science and Engineering ， Harbin Institute of technology ， Harbin 150001 ， China ） （ 2 . Institute of Engineering Thermophysics ， Academia Sinica ， Beijing 100080 ， China ） Abstract ： In order to improve turbine internaI efficiency and Iower manufacturing cost ， a new highIy Ioaded rotating bIade has been deveIoped. The 3D optimization design method based on artificiaI neuraI network and genetic aIgorithm is adopted to construct the bIade shape. The bIade is stacked by the center of gravity in radiaI direction with five sections. For each bIade section ， independent suction and pressure sides are constructed from the camber Iine using Bezier curves. Three-dimensionaI fIow anaIysis is carried out to verify the performance of the new bIade. It is found that the new bIade has improved the bIade performance by 0.5% . ConseguentIy ， it is verified that the new bIade is effective to improve the turbine internaI efficiency and to Iower the turbine weight and manufacturing cost by reducing the bIade number by about 15% . Key words ： optimization design ； highIy Ioaded ； rotating bIades ； artificiaI neuraI network ； genetic aIgorithm 采用基于神经网络及遗传算法的叶轮机械叶片三维优化设计方法开发高载荷透平动叶片 . 周凡 贞，冯国泰，蒋洪德 . 中国航空学报（英文版）， 2003 ， 16 （ 4 ）： 198 - 202. 摘 要： 为了提高透平的内效率及降低制造成本，采用基于神经网络及遗传算法的叶轮机械叶 片三维优化设计方法，开发了一种高载荷动叶片。该叶片是以 5 个截面的重心沿径向积叠而成。 每个截面的内背弧是以基于叶型中弧线的 Bezier 曲线来构造。三维流动分析表明，新叶片提高效 率 0.5% 。通过减少叶片数约 15% ，新开发的高载荷动叶片不仅有效地提高了透平的内效率，同时 降低了透平重量和制造成本。 关键词： 优化设计；高载荷；透平动叶片；人工神经网络；遗传算法 文章编号： 1000-9361 （ 2003 ） 04-0198-05 中图分类号： V235.1 文献标识码： A In the fieId of turbomachine