固体力学与飞行器总体设计

样本分簇增强的快速近似建模方法及应用

  • 王悦 ,
  • 王志祥 ,
  • 李道奎 ,
  • 雷勇军
展开
  • 1.国防科技大学 空天科学学院,长沙 410073
    2.空天任务智能规划与仿真湖南省重点实验室,长沙 410073
    3.国防科技大学 高超声速技术实验室,长沙 410073
    4.火箭军工程大学,西安 710025

收稿日期: 2025-01-02

  修回日期: 2025-02-10

  录用日期: 2025-04-08

  网络出版日期: 2025-04-17

基金资助

国家自然科学基金(52405301);湖南省自然科学基金(2024JJ6453)

A fast approximate modeling method and application for sample cluster enhancement

  • Yue WANG ,
  • Zhixiang WANG ,
  • Daokui LI ,
  • Yongjun LEI
Expand
  • 1.College of Aerospace Science and Engineering,National University of Defense Technology,Changsha 410073,China
    2.Hunan Key Laboratory of Intelligent Planning and Simulation for Aerospace Missions,Changsha 410073,China
    3.Hypersonic Technology Laboratory,National University of Defense Technology,Changsha 410073,China
    4.Rocket Force University of Engineering,Xi’an 710025,China

Received date: 2025-01-02

  Revised date: 2025-02-10

  Accepted date: 2025-04-08

  Online published: 2025-04-17

Supported by

National Natural Science Foundation of China(52405301);Natural Science Foundation of Hunan Province(2024JJ6453)

摘要

为提高复合材料加筋圆柱壳后屈曲分析和优化效率,提出了一种融合样本分簇和改进K折交叉验证的增广径向基函数(ARBF)快速近似建模方法。采用K-means聚类算法确定了样本最优分簇,基于样本局部密度确定了样本的基准形状参数,各分簇均引入缩放系数自适应调整形状参数,有效兼顾了优化形状参数的效率和精度。通过样本子集建立ARBF辅助近似模型,进而建立了基于偏差-方差分解的ARBF辅助近似模型泛化性能评估准则,解决了传统K折交叉验证样本信息利用不足的难题;基于分块矩阵求逆技术推导了ARBF辅助近似模型的高阶系数矩阵快速求逆方法,提出了基于改进K折交叉验证的缩放系数优化方法,大幅降低了确定最优形状参数的计算复杂度,提升了ARBF近似建模效率和精度。数值和工程算例表明,样本最优分簇和快速交叉验证对近似建模效率和精度有显著增益,降低了建模效率对样本规模和问题维度的敏感性,且相同训练样本数量下,该方法建模精度显著优于其他典型方法,验证了该方法的有效性和先进性,具有一定的工程应用价值。

本文引用格式

王悦 , 王志祥 , 李道奎 , 雷勇军 . 样本分簇增强的快速近似建模方法及应用[J]. 航空学报, 2025 , 46(22) : 231755 -231755 . DOI: 10.7527/S1000-6893.2025.31755

Abstract

To improve the analysis and optimization efficiency of stiffened composite cylindrical shells, a novel fast modeling method for the Augmented Radial Basis Function (ARBF) approximate model is proposed by fusing sample clustering and improved K-fold cross-validation. The K-means clustering algorithm is used to determine the optimal sample clustering, and the reference shape parameters of the samples are determined based on the local density of the samples. The scaling coefficient is introduced into each cluster to adjust the shape parameters adaptively, which effectively balance the optimization efficiency and accuracy of the shape parameters. The ARBF auxiliary approximation model is established by sample subset, and then the evaluation criteria of ARBF auxiliary approximation model generalization performance is established based on bias-variance decomposition, addressing the problem of insufficient utilization of traditional K-fold cross-validation sample information. Based on the block matrix inversion technique, a fast inversion method of the high-order coefficient matrix of the ARBF auxiliary approximation model is derived, and the scaling coefficient optimization method based on the improved K-fold cross-validation is proposed, which greatly reduces the computational complexity of determining the optimal shape parameters and improves the efficiency and accuracy of ARBF approximation modeling. Numerical and engineering examples validate that the optimal sample clustering and fast cross-validation have significant contributions to the approximate modeling efficiency and accuracy, and reduce the sensitivity of modeling efficiency to sample size and problem dimension. Moreover, using the same number of training samples, the modeling accuracy of the proposed method is significantly superior to that of other typical methods. The results verifies the effectiveness, advancemen and practical engineering value of the proposed method.

参考文献

[1] 熊健, 李志彬, 刘惠彬, 等. 航空航天轻质复合材料壳体结构研究进展[J]. 复合材料学报202138(6): 1629-1650.
  XIONG J, LI Z B, LIU H B, et al. Advances in aerospace lightweight composite shell structure[J]. Acta Materiae Compositae Sinica202138(6): 1629-1650 (in Chinese).
[2] 王博, 郝鹏, 田阔. 加筋薄壳结构分析与优化设计研究进展[J]. 计算力学学报201936(1): 1-12.
  WANG B, HAO P, TIAN K. Recent advances in structural analysis and optimization of stiffened shells[J]. Chinese Journal of Computational Mechanics201936(1): 1-12 (in Chinese).
[3] LI M, ZHU H Y, LAI C L, et al. Recent progresses in lightweight carbon fibre reinforced lattice cylindrical shells[J]. Progress in Aerospace Sciences2022135: 100860.
[4] 张超, 曹勇, 赵振强, 等. 树脂基复合材料在民用航空发动机中的应用与关键技术研究进展[J]. 航空学报202445(2): 028556.
  ZHANG C, CAO Y, ZHAO Z Q, et al. Applications and key challenges of polymer composites in civil aero-engines: State-of-art review[J]. Acta Aeronautica et Astronautica Sinica202445(2): 028556 (in Chinese).
[5] ZHANG L, HE W P, ZHU S Y, et al. Hierarchical lattice stiffened cylindrical shell: Design, high-order vibration theory and composite parameter identification[J]. Composite Structures2024328: 117725.
[6] YAZDANI M, SHARIFLOO S, REZAEI M. Buckling behavior and energy absorption of grid-stiffened composite cylinders under low-velocity axial impact[J]. Polymer Composites202445(3): 2474-2484.
[7] 陈向明, 李新祥, 柴亚南, 等. 复合材料壁板后屈曲设计与分析技术研究进展[J]. 复合材料学报202441(9): 4673-4700.
  CHEN X M, LI X X, CHAI Y N, et al. Research progress in post-buckling design and analysis techniques for composite stiffened panel[J]. Acta Materiae Compositae Sinica202441(9): 4673-4700 (in Chinese).
[8] LI M, ZHANG L, TAO Y M, et al. Two-level homogenization and multi-failure analyses of hierarchical lattice stiffened cylinder[J]. Composite Structures2023323: 117487.
[9] ZHANG L W, ZHAO L Y, PAN L, et al. Research progress of composite cylindrical shells[J]. Polymer Composites202344(11): 7298-7316.
[10] 张志峰, 陈浩然, 李煊, 等. 先进复合材料格栅圆柱壳优化设计的混合遗传算法[J]. 复合材料学报200522(2): 166-171.
  ZHANG Z F, CHEN H R, LI X, et al. Hybrid genetic algorithm for optimum design of advanced grid composite circular cylinders[J]. Acta Materiae Compositae Sinica200522(2): 166-171.
[11] 王文煜, 李锋, 任飞翔, 等. 轻质高强复合材料网格加筋壳体结构设计方法及力学性能研究进展[J]. 航空学报202445(17): 530001.
  WANG W Y, LI F, REN F X, et al. Research progress on structural design methods and mechanical properties of lightweight high-strength composite lattice stiffened shell structure[J]. Acta Aeronautica et Astronautica Sinica202445(17): 530001 (in Chinese).
[12] 李增聪, 田阔, 黄蕾, 等. 面向变刚度复合材料筒壳高效屈曲分析的变保真度迁移学习模型[J]. 复合材料学报202239(5): 2430-2440.
  LI Z C, TIAN K, HUANG L, et al. Variable-fidelity transfer learning model for efficient buckling analysis of variable stiffness composite cylindrical shells[J]. Acta Materiae Compositae Sinica202239(5): 2430-2440 (in Chinese).
[13] 韩忠华, 许晨舟, 乔建领, 等. 基于代理模型的高效全局气动优化设计方法研究进展[J]. 航空学报202041(5): 623344.
  HAN Z H, XU C Z, QIAO J L, et al. Recent progress of efficient global aerodynamic shape optimization using surrogate-based approach[J]. Acta Aeronautica et Astronautica Sinica202041(5): 623344 (in Chinese).
[14] 文谦, 杨家伟, 武泽平, 等. 快速交叉验证改进的运载火箭近似建模方法[J]. 航空学报202243(9): 225967.
  WEN Q, YANG J W, WU Z P, et al. An approximation modeling method of launch vehicles improved by fast cross-validation[J]. Acta Aeronautica et Astronautica Sinica202243(9): 225967 (in Chinese).
[15] 胡嘉欣, 芮姝, 高瑞朝, 等. 飞行器结构布局与尺寸混合优化方法[J]. 航空学报202243(5): 225363.
  HU J X, RUI S, GAO R C, et al. Hybrid optimization method for structural layout and size of flight vehicles[J]. Acta Aeronautica et Astronautica Sinica202243(5): 225363 (in Chinese).
[16] CAVORETTO R, DE ROSSI A, MUKHAMETZHANOV M S, et al. On the search of the shape parameter in radial basis functions using univariate global optimization methods[J]. Journal of Global Optimization202179(2): 305-327.
[17] 谢志超. 高斯径向基函数形状参数选取方法研究[D]. 唐山: 华北理工大学, 2023.
  XIE Z C. Study on shape parameter selection method of Gaussian radial basis function[D]. Tangshan: North China University of Science and Technology, 2023 (in Chinese).
[18] DU W Y, MA J, YUE P, et al. An efficient reliability method with multiple shape parameters based on radial basis function[J]. Applied Sciences202212(19): 9689.
[19] CHEN C S, NOORIZADEGAN A, YOUNG D L, et al. On the selection of a better radial basis function and its shape parameter in interpolation problems[J]. Applied Mathematics and Computation2023442: 127713.
[20] KITAYAMA S, YAMAZAKI K. Simple estimate of the width in Gaussian kernel with adaptive scaling technique[J]. Applied Soft Computing201111(8): 4726-4737.
[21] YAO W, CHEN X Q, ZHAO Y, et al. Concurrent subspace width optimization method for RBF neural network modeling[J]. IEEE Transactions on Neural Networks and Learning Systems201223(2): 247-259.
[22] ZHANG Y W, GONG C L, FANG H, et al. An efficient space division-based width optimization method for RBF network using fuzzy clustering algorithms[J]. Structural and Multidisciplinary Optimization201960(2): 461-480.
[23] YANG C, WANG Z, OH S K, et al. Ensemble fuzzy radial basis function neural networks architecture driven with the aid of multi-optimization through clustering techniques and polynomial-based learning[J]. Fuzzy Sets and Systems2022438: 62-83.
[24] WU Z P, WANG D H, PATRICK OKOLO N, et al. Unified estimate of Gaussian kernel width for surrogate models[J]. Neurocomputing2016203: 41-51.
[25] YANG J W, WU Z P, WANG Z X, et al. Enhanced anisotropic radius basis function metamodel based on recursive evolution Latin hypercube design and fast K-fold cross-validation[J]. Structural and Multidisciplinary Optimization202366(7): 169.
[26] 王志祥, 武泽平, 王婕, 等. 大型运载火箭加筋柱壳近似建模方法[J]. 宇航学报202041(10): 1267-1279.
  WANG Z X, WU Z P, WANG J, et al. Approximation modeling method for cylindrical stiffened shells in large launch vehicles[J]. Journal of Astronautics202041(10): 1267-1279 (in Chinese).
[27] 张杰, 李国盛, 文谦, 等. 基于改进增广径向基的固体姿轨控发动机推力快速预示[J]. 推进技术202344(8): 206-214.
  ZHANG J, LI G S, WEN Q, et al. Fast thrust prediction method for solid divert and attitude control system based on improved augmented radial basis functions[J]. Journal of Propulsion Technology202344(8): 206-214 (in Chinese).
[28] ARLOT S, CELISSE A. A survey of cross-validation procedures for model selection[J]. Statistics Surveys20104: 40-79.
[29] LI H X, WANG X L. Cross validation based model averaging for varying-coefficient models with response missing at random[J]. Journal of Applied Mathematics and Physics202412(3): 764-777.
[30] LILAND K H, SKOGHOLT J, INDAHL U G. A new formula for faster computation of the K-fold cross-validation and good regularisation parameter values in ridge regression[J]. IEEE Access202412: 17349-17368.
[31] 王志祥. 重型运载火箭加筋圆柱壳舱段结构优化与试验[D]. 长沙: 国防科技大学, 2021.
  WANG Z X. Optimization design and experiment of stiffened cylindrical shells in heavy-lift launch vehicle[D]. Changsha: National University of Defense Technology, 2021 (in Chinese).
[32] ZHOU X J, MA Y Z, LI X F. Ensemble of surrogates with recursive arithmetic average[J]. Structural and Multidisciplinary Optimization201144(5): 651-671.
[33] JIANG G X, WANG W J. Error estimation based on variance analysis of K-fold cross-validation[J]. Pattern Recognition201769: 94-106.
[34] YE Y F, WANG Z X, ZHANG X B. An optimal pointwise weighted ensemble of surrogates based on minimization of local mean square error[J]. Structural and Multidisciplinary Optimization202062(2): 529-542.
[35] LI J. Asymptotics of K-fold cross validation[J]. Journal of Artificial Intelligence Research202378: 491-526.
[36] HASTIE T, FRIEDMAN J, TIBSHIRANI R. The elements of statistical learning[M]. New York: Springer New York, 2001
[37] 武泽平. 基于数值模拟的序列近似优化方法研究[D]. 长沙: 国防科技大学, 2018.
  WU Z P. Study on numerical simulation based sequential approximate optimization[D]. Changsha: National University of Defense Technology, 2018 (in Chinese).
[38] WANG Z X, LEI Y J, CUI H R, et al. Enhanced RBF neural network metamodelling approach assisted by sliced splitting-based K-fold cross-validation and its application for the stiffened cylindrical shells[J]. Aerospace Science and Technology2022124: 107534.
[39] 杨俊闯, 赵超. K-means聚类算法研究综述[J]. 计算机工程与应用201955(23): 7-14, 63.
  YANG J C, ZHAO C. Survey on K-means clustering algorithm[J]. Computer Engineering and Applications201955(23): 7-14, 63 (in Chinese).
[40] EZUGWU A E, IKOTUN A M, OYELADE O O, et al. A comprehensive survey of clustering algorithms: State-of-the-art machine learning applications, taxonomy, challenges, and future research prospects[J]. Engineering Applications of Artificial Intelligence2022110: 104743.
[41] SHI C M, WEI B T, WEI S L, et al. A quantitative discriminant method of elbow point for the optimal number of clusters in clustering algorithm[J]. EURASIP Journal on Wireless Communications and Networking20212021(1): 31.
[42] 王建仁, 马鑫, 段刚龙. 改进的K-means聚类k值选择算法[J]. 计算机工程与应用201955(8): 27-33.
  WANG J R, MA X, DUAN G L. Improved K-means clustering k-value selection algorithm[J]. Computer Engineering and Applications201955(8): 27-33 (in Chinese).
[43] SELVIDA D, ZARLIS M, SITUMORANG Z. Analysis of the effect early cluster centre points on the combination of K-means algorithms and sum of squared error on k centroid[J]. IOP Conference Series: Materials Science and Engineering2020725(1): 012089.
[44] KOHAVI R. A study of cross-validation and bootstrap for accuracy estimation and model selection[C]∥Proceedings of the 14th International Joint Conference on Artificial Intelligence-Volume 2. New York: ACM, 1995: 1137-1143.
[45] 李正良, 彭思思, 王涛. 基于K-fold交叉验证的代理模型序列采样方法[J]. 计算力学学报202239(2): 244-249.
  LI Z L, PENG S S, WANG T. A sequential sampling method of surrogate model based on K-fold cross validation[J]. Chinese Journal of Computational Mechanics202239(2): 244-249 (in Chinese).
[46] LUO J C, CHEN Q X, ZHENG J, et al. An elliptical basis function network for classification of remote sensing images[J]. Journal of Geographical Systems20046(3): 219-236.
[47] SIMPSON T W, MAUERY T M, KORTE J J, et al. Kriging models for global approximation in simulation-based multidisciplinary design optimization[J]. AIAA Journal200139(12): 2233-2241.
[48] ROUSSEEUW P J. Silhouettes: A graphical aid to the interpretation and validation of cluster analysis[J]. Journal of Computational and Applied Mathematics198720: 53-65.
[49] BAGIROV A M, ALIGULIYEV R M, SULTANOVA N. Finding compact and well-separated clusters: Clustering using silhouette coefficients[J]. Pattern Recognition2023135: 109144.
[50] LI M, FAN H L. Multi-failure analysis of composite Isogrid stiffened cylinders[J]. Composites Part A: Applied Science and Manufacturing2018107: 248-259.
[51] HASHIN Z. Fatigue failure criteria for unidirectional fiber composites[J]. Journal of Applied Mechanics198148(4): 846-852.
[52] 耿发贵, 李强, 宋薛思, 等. 基于冲击损伤的复合材料气瓶铺层顺序优化设计[J]. 复合材料学报202239(2): 777-787.
  GENG F G, LI Q, SONG X S, et al. Optimal design of laying sequence of composite gas cylinders based on impact damage[J]. Acta Materiae Compositae Sinica202239(2): 777-787 (in Chinese).
文章导航

/