首页 >

基于自适应采样的高超声速飞行器气动热全局快速预示

杨国涛,岳振江,刘莉   

  1. 北京理工大学
  • 收稿日期:2022-05-09 修回日期:2022-09-27 出版日期:2022-09-30 发布日期:2022-09-30
  • 通讯作者: 岳振江

Rapid Predicting of Global Hypersonic Vehicle Aerothermodynamics Based on Adaptive Sampling

Guo-Tao YANG1,Zhen-Jiang YUELi LIU   

  • Received:2022-05-09 Revised:2022-09-27 Online:2022-09-30 Published:2022-09-30
  • Contact: Zhen-Jiang YUE

摘要: 高超声速飞行器热防护系统设计中高精度气动热分析模型使得设计计算成本不断增加,基于数据驱动的气动热环境预示方法受到广泛关注。针对有限高精度模型计算成本下提升全局预示精度的问题,提出一种基于模糊聚类的批量自适应采样方法,根据预示误差分布特征通过聚类采用超球分割构建采样影响域,兼顾误差较大的重点采样域与全局探索;通过当地误差评分系数加权构建采样拒绝域,减小新增样本冗余,结合maxmin准则在综合确定的重点采样空间中新增样本,提升采样质量,进而实现预示模型全局精度快速提升。数值测试算例表明,所提方法与One-Shot、APSFC、CV–Voronoi方法相比能有效降低所需采样规模,加速提升预示精度。通过类HTV-2飞行器气动热快速预示实例,验证了方法的有效性与工程实用性。

关键词: 自适应采样, 模糊聚类, 快速预示, 高超声速, 气动热

Abstract: High-fidelity aerothermodynamics analysis models in the thermal protection system of the hypersonic vehicles sig-nificantly increase the computational budget of engineering design, thus rapid predicting methods based on data-driven have been widely concerned recently. In this paper, a batch adaptive sampling method based on fuzzy clus-tering is proposed, aiming at improving the global prediction accuracy with the limited computational budget of high fidelity models. The sampling influence domain is constructed by clustering and hypersphere segmentation under the distribution characteristics of the predicting error, which takes into account the key sampling domain with larger error and exploration. The sampling refused domain is constructed by local error scoring coefficient weighted to reduce the redundancy of newly adding samples. The method adds new samples in the comprehensively deter-mined key sampling space to improve the sampling quality based on maxmin criterion, thereby the global accuracy of the predicting models have been improved rapidly. The comparison results show that the proposed method out-performs One-Shot, APSFC and CV–Voronoi in terms of reducing the sampling scale required and speeding up to improve predicting accuracy. The rapid predicting results of HTV-2 typed vehicle aerothermodynamics demon-strates the practicality and effectiveness of the proposed method in engineering practices.

Key words: adaptive sampling, fuzzy clustering, rapid predicting, hypersonic, aerothermodynamics

中图分类号: