航空学报 > 2011, Vol. 32 Issue (1): 99-106   doi: CNKI:11-1929/V.20101111.0909.001

基于Kriging方法的航空发动机压气机特性元建模

尹大伟1, 李本威2, 王永华2, 冯向军1   

  1. 1. 国防科学技术大学 航天与材料工程学院, 湖南 长沙 410073;2. 海军航空工程学院 飞行器工程系, 山东 烟台 264001
  • 收稿日期:2010-04-25 修回日期:2010-06-08 出版日期:2011-01-25 发布日期:2011-01-25

Aeroengine Compressor Characteristics Metamodeling Using Kriging Method

YIN Dawei1, LI Benwei2, WANG Yonghua2, FENG Xiangjun1   

  1. 1. College of Aerospace and Material Engineering, National University of Defense Technology, Changsha 410073, China;2. Department of Aerocraft Engineering, Naval Aeronautical and Astronautical Academy, Yantai 264001, China
  • Received:2010-04-25 Revised:2010-06-08 Online:2011-01-25 Published:2011-01-25

摘要: 针对求解航空发动机数学模型时,使用传统的压气机二维特性曲线进行插值计算精度较差的问题,引入了优化技术中元建模的思想,对压气机特性进行模型重构。根据压气机特性数据空间分布的特点,定义元模型的输入输出维数,采用Kriging算法建立了流量和效率特性的代理模型。以某型涡扇发动机为例进行了压气机特性元建模的仿真计算,设定了满足压气机稳定工作要求的计算区域,在新的计算区域上对未知转速下的特性进行了模型重构。通过模型的校核与验证表明,根据Kriging代理模型重构的部件特性有较高的精度,优于传统的插值方法。可以预期压气机特性Kriging代理模型的应用,对提高航空发动机模型求解的精度,具有一定的工程应用价值。

关键词: 航空发动机, Kriging方法, 元建模, 压气机特性, 代理模型

Abstract: A metamodeling method used in optimization is introduced to reconstruct the characteristics models of an aeroengine compressor, because the traditional 2D compressor characteristics maps used in the interpolation for aeroengine model solving lack sufficient precision. According to the space distribution of the compressor characteristics data, the dimensions of the metamodel inputs and outputs are defined, and the flow and efficiency surrogate models are established using the Kriging method. The simulation on a given turbofan engine is performed as an example, in which the calculation region is set to satisfy its steady work, and then the characteristics of unknown rotational speeds are reconstructed in the new calculation region based on the established Kriging metamodels. The results of model verification and validation show that, the reconstructed characteristics using Kriging surrogate models have high precision, and are better than the linear interpolation method. It can be predicted that the application of compressor characteristics Kriging metamodels will improve the precision of the aeroengine model solution. Therefore, the study is of significant engineering value.

Key words: aeroengine, Kriging method, metamodeling, compressor characteristics, surrogate model

中图分类号: