基于主动学习Kriging的飞行器射程评估  

Vehicle Range Assessment Based on Active Learning Kriging

作　　者：王浩[1] 张海瑞[1] 王尧 洪东跑[1] 何沛昊 WANG Hao;ZHANG Hairui;WANG Yao;HONG Dongpao;HE Peihao(China Academy of Launch Vehicle Technology,Beijing 100076,China)

机构地区：[1]中国运载火箭技术研究院,北京100076

出　　处：《兵器装备工程学报》2019年第5期56-60,共5页Journal of Ordnance Equipment Engineering

基　　金：军委装备发展部"十三五"装备预研领域基金项目(6140244010216HT15001)

摘　　要：提出了一种基于主动学习Kriging的飞行器射程评估方法;结合高超声速飞行器的特点,通过将气动、推进、质量学科的不确定性因素注入飞行器设计多学科分析模型,构建飞行器射程概率评估模型;进而,利用主动学习策略序列加点方法建立射程评估代理模型,实现对射程及设计裕度的量化分析;仿真结果表明:该方法合理可行,显著提升了射程评估效率。Vehicle range assessment based on active learning Kriging was proposed to improve the vehicle range quantization accuracy.In this method,the uncertainties of aerodynamic/propellant/mass model were injected directly into the multidisciplinary analysis model.And the range probability assessment model was constructed.The active learning Kriging was used to add the training point sequentially and update the Kriging model adaptively until convergence.And the range reliability assessment can be obtained efficiently.It is shown that the method is feasible and the efficiency of the range assessment is improved enormously.

关 键 词：高超声速飞行器 射程评估 裕度量化 主动学习 KRIGING模型 

分 类 号：V417.2[航空宇航科学与技术—航空宇航推进理论与工程] O213.2[理学—概率论与数理统计]