基于随机森林的飞行载荷代理模型分析方法

doi:10.7527/S1000-6893.2021.25640

固体力学与飞行器总体设计

本期目录 | 过刊浏览 | 高级检索

前一篇 | 后一篇

基于随机森林的飞行载荷代理模型分析方法

李海泉^1,2, 陈小前^1,3, 左林玄², 赵卓林²

1. 国防科技大学空天科学学院, 长沙 410073;
2. 中国航空工业集团有限公司沈阳飞机设计研究所, 沈阳 110035;
3. 中国人民解放军军事科学院国防科技创新研究院, 北京 100071

收稿日期:2021-05-14 修回日期:2021-05-14 发布日期:2021-05-31
通讯作者: 李海泉 E-mail:haiquan.li601@163.com
基金资助:
国家自然科学基金（61903349）

Surrogate model for flight load analysis based on random forest

LI Haiquan^1,2, CHEN Xiaoqian^1,3, ZUO Linxuan², ZHAO Zhuolin²

1. College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China;
2. Shenyang Aircraft Design Institute, Aviation Industry Corporation of China, Shenyang 110035, China;
3. National Innovation Institute of Defense Technology, Chinese Academy of Military Science, Beijing 100071, China

Received:2021-05-14 Revised:2021-05-14 Published:2021-05-31
Supported by:
National Natural Science Foundation of China (61903349)

摘要/Abstract

摘要： 飞行载荷分析（FLC）通常需消耗大量的计算资源和时间，提升飞行载荷计算效率对于缩短研发周期、提高设计质量具有重要意义。本文研究了数据驱动的机器学习代理模型飞行载荷分析方法，由于随机森林（RF）代理模型具有学习效率较高、泛化能力较强、可避免过拟合、参数可解释、变量敏感度分析等优点，使其十分切合飞行载荷分析，具有重要的应用潜力和前景。文中基于传统的飞行载荷分析方法，采用NASTRAN的SQL144载荷分析框架获得训练随机森林代理模型的样本数据，然后以高度、马赫数、过载、俯仰角加速度等作为输入参数构建了飞机对称机动载荷预测代理模型。采用建立的模型预测了算例飞机其他工况机翼翼根、平尾翼根的剪力、弯矩、扭矩，通过对预测结果评估校验，证实了模型具有较高的精度，可以大幅提升飞行载荷分析效率，并能够分析飞行载荷对各状态参数的敏感度，为高效全面地分析飞行载荷提供了新的思路。

关键词: 飞行载荷, 代理模型, 机器学习, 随机森林, 对称机动

Abstract: Flight Load Computation (FLC) consumes considerable computational time and resources, and promoting FLC efficiency is highly significant to shorten the development cycle and improve the design quality of aircraft. This paper focuses on a data-driven machine learning surrogate model for load analysis based on the Random Forest (RF), which fits the load analysis with important potential and application prospect in this field because of its advantages such as high learning efficiency, strong generalization ability, overfitting avoidance, parameter interpretability, and variable sensitivity analysis. The training data for the RF surrogate model using conventional load analysis methods and the SQL144 framework of NASTRAN are generated, and parameters such as height, Mach number, overload, and pitch angular acceleration serve as the input variables to establish the surrogate model of load prediction in the case of symmetric maneuver. The proposed model is used to predict other conditions and the shear force, bending moment, and torque of roots of wings and horizontal tail. The predicting results are evaluated and verified, demonstrating high accuracy of the RF surrogate model and its ability to drastically promote the FLC efficiency and conduct sensitivity analysis of the flight load to state parameters. This study provides a new approach to the efficient and comprehensive flight load analysis.

Key words: flight load, surrogate model, machine learning, random forest, symmetric maneuver

中图分类号:

V215.1

李海泉, 陈小前, 左林玄, 赵卓林. 基于随机森林的飞行载荷代理模型分析方法[J]. 航空学报, 2022, 43(3): 225640-225640.

LI Haiquan, CHEN Xiaoqian, ZUO Linxuan, ZHAO Zhuolin. Surrogate model for flight load analysis based on random forest[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2022, 43(3): 225640-225640.

参考文献

[1] MOHAMED K S. Machine learning for model order reduction[M]. Berlin:Springer International Publishing, 2018.
[2] GOODFELLOW I, BENGIO Y, COURVILLE A. Deep learning[M]. Cambridge:MIT Press, 2016:3-6.
[3] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Piscataway:IEEE Press, 2016:770-778.
[4] KUTZ J N. Deep learning in fluid dynamics[J]. Journal of Fluid Mechanics, 2017, 814:1-4.
[5] LING J L, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance[J]. Journal of Fluid Mechanics, 2016, 807:155-166.
[6] 张天姣, 钱炜祺, 周宇, 等. 人工智能与空气动力学结合的初步思考[J]. 航空工程进展, 2019, 10(1):1-11. ZHANG T J, QIAN W Q, ZHOU Y, et al. Preliminary thoughts on the combination of artificial intelligence and aerodynamics[J]. Advances in Aeronautical Science and Engineering, 2019, 10(1):1-11(in Chinese).
[7] 王博斌, 张伟伟, 叶正寅. 基于神经网络模型的动态非线性气动力辨识方法[J]. 航空学报, 2010, 31(7):1379-1388. WANG B B, ZHANG W W, YE Z Y. Unsteady nonlinear aerodynamics identification based on neural network model[J]. Acta Aeronautica et Astronautica Sinica, 2010, 31(7):1379-1388(in Chinese).
[8] 叶舒然, 张珍, 王一伟, 等. 基于卷积神经网络的深度学习流场特征识别及应用进展[J]. 航空学报, 2021, 42(4):524736. YE S R, ZHANG Z, WANG Y W, et al. Progress in deep convolutional neuralnetwork based flow field recognition and its applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524736(in Chinese).
[9] SUN Z W, WANG C, ZHENG Y, et al. Non-intrusive reduced-order model for predicting transonic flow with varying geometries[J]. Chinese Journal of Aeronautics, 2020, 33(2):508-519.
[10] GOU Y Y, LI H B, DONG X M, et al. Constrained adaptive neural network control of an MIMO aeroelastic system with input nonlinearities[J]. Chinese Journal of Aeronautics, 2017, 30(2):796-806.
[11] SECCO N R, DE MATTOS B S. Artificial neural networks to predict aerodynamic coefficients of transport airplanes[J]. Aircraft Engineering and Aerospace Technology, 2017, 89(2):211-230.
[12] 史志伟, 明晓. 基于模糊聚类的模糊神经网络在非定常气动力建模中的应用[J]. 空气动力学学报, 2005, 23(1):21-24. SHI Z W, MING X. The application of FNN in unsteady aerodynamics modeling based on fuzzy clustering[J]. Acta Aerodynamica Sinica, 2005, 23(1):21-24(in Chinese).
[13] 何磊, 钱炜祺, 汪清, 等. 机器学习方法在气动特性建模中的应用[J]. 空气动力学学报, 2019, 37(3):470-479. HE L, QIAN W Q, WANG Q, et al. Applications of machine learning for aerodynamic characteristics modeling[J]. Acta Aerodynamica Sinica, 2019, 37(3):470-479(in Chinese).
[14] WANG Q, QIAN W Q, HE K F. Unsteady aerodynamic modeling at high angles of attack using support vector machines[J]. Chinese Journal of Aeronautics, 2015, 28(3):659-668.
[15] IGNATYEV D, KHRABROV A. Experimental study and neural network modeling of aerodynamic characteristics of canard aircraft at high angles of attack[J]. Aerospace, 2018, 5(1):26.
[16] IGNATYEV D I, KHRABROV A N. Neural network modeling of unsteady aerodynamic characteristics at high angles of attack[J]. Aerospace Science and Technology, 2015, 41:106-115.
[17] LING J L, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance[J]. Journal of Fluid Mechanics, 2016, 807:155-166.
[18] HOLMES G, SARTOR P, REED S, et al. Prediction of landing gear loads using machine learning techniques[J]. Structural Health Monitoring, 2016, 15(5):568-582.
[19] JEONG S H, LEE K B, HAM J H, et al. Estimation of maximum strains and loads in aircraft landing using artificial neural network[J]. International Journal of Aeronautical and Space Sciences, 2020, 21(1):117-132.
[20] HALLE M, THIELECKE F. Local Model Networks applied to Flight Loads Estimation[C]//31st Congress of the International Council of the Aeronautical Sciences (ICAS), 2018.
[21] ALLEN M J, DIBLEY R P. Modeling aircraft wing loads from flight data using neural networks[C]//SAE Technical Paper Series. 400 Commonwealth Drive. Warrendale:SAE International, 2003.
[22] COOPER S B, DIMAIO D. Static load estimation using artificial neural network:Application on a wing rib[J]. Advances in Engineering Software, 2018, 125:113-125.
[23] 赵永辉. 气动弹性力学与控制[M]. 北京:科学出版社, 2007:267-283. ZHAO Y H. Aeroelasticity and control[M]. Beijing:Science Press, 2007:267-283(in Chinese).
[24] BREIMAN L. Random forests[J]. Machine Learning, 2001, 45:5-32.

E-mail：hkxb@buaa.edu.cn

关于我们

期刊社服务

专业学科

封面文章

友情链接

主管单位：中国科学技术协会主办单位：中国航空学会北京航空航天大学

基于随机森林的飞行载荷代理模型分析方法

Surrogate model for flight load analysis based on random forest

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	崔壮壮, 原昕, 赵国庆, 井思梦, 招启军. 共轴刚性旋翼高速直升机前飞性能操纵策略影响[J]. 航空学报, 2024, 45(9): 529256-529256.
[2]	柳家齐, 陈荣钱, 楼锦华, 韩旭, 吴昊, 尤延铖. 基于深度学习的高速直升机旋翼翼型气动优化设计[J]. 航空学报, 2024, 45(9): 529828-529828.
[3]	李玉涵, 杨宝玉, 吴亦农, 张强, 唐晓. 卫星光机载荷热模型参数高效修正方法研究进展[J]. 航空学报, 2024, 45(6): 628814-628814.
[4]	陈树生, 冯聪, 张兆康, 赵轲, 张新洋, 高正红. 基于全局/梯度优化方法的宽速域乘波-机翼布局气动设计[J]. 航空学报, 2024, 45(6): 629596-629596.
[5]	马金毅, 王灿, 薛涛, 艾剑良, 董一群. 空战格斗飞行机动数据库建立及应用[J]. 航空学报, 2023, 44(S1): 727538-727538.
[6]	赵欢, 高正红, 夏露. 基于新型多可信度代理模型的多目标优化方法[J]. 航空学报, 2023, 44(6): 126962-126962.
[7]	何磊, 钱炜祺, 董康生, 易贤, 柴聪聪. 基于卷积神经网络的结冰翼型气动特性建模[J]. 航空学报, 2023, 44(5): 126434-126434.
[8]	赵欢, 高正红, 夏露. 基于新型高维代理模型的气动外形设计方法[J]. 航空学报, 2023, 44(5): 126924-126924.
[9]	陈唯实, 刘佳, 王青斌, 卢贤锋, 张洁, 陈小龙, 黄毅峰. 气象雷达探鸟技术综述[J]. 航空学报, 2023, 44(5): 26781-026781.
[10]	郑志阳, 张阳, 张钊, 吴宝海, 张莹. 基于GA⁃SVR的薄壁叶片辅助支撑布局优化方法[J]. 航空学报, 2023, 44(4): 426805-426805.
[11]	于汀, 李璐祎, 刘昱杉, 常泽明. 观测不确定性下的高效贝叶斯更新方法及其在机翼结构中的应用[J]. 航空学报, 2023, 44(24): 228592-228592.
[12]	余建虎. 飞机飞行载荷校准试验约束补偿技术[J]. 航空学报, 2023, 44(19): 227500-227500.
[13]	梁益铭, 李广宁, 徐敏. 基于机器学习的智能控制数值虚拟飞行方法[J]. 航空学报, 2023, 44(17): 128098-81280986.
[14]	陈小前, 赵勇, 霍森林, 张泽雨, 都柄晓. 多尺度结构拓扑优化设计方法综述[J]. 航空学报, 2023, 44(15): 528863-528863.
[15]	员婉莹, 吕震宙. 参数可靠性全局灵敏度高效分析的代理模型法[J]. 航空学报, 2023, 44(12): 227670-227670.