目的 基于传统更新方法常受制于效率低、自动化程度有限以及依赖人为主观决策等问题,提出一种基于深度学习的三阶段分类框架,实现面向导管架结构的智能端到端有限元模型修正。方法 将“定性判别-定位-定量（QLQ）”的决策流程具体化,定性判别是否需要更新,定位需更新的区域,定量评估更新幅度。通过共享特征提取器与双重注意力设计的结合,可直接从原始结构响应数据中学习判别性表征,从而自动化完成完整的更新流程。结果 基于包含83 421个载荷工况的数据集（训练/验证/测试划分为7︰1.5︰1.5）,所提方法在各阶段均取得稳定而优异的性能。更新检测准确率为97.21%,区域分类准确率为93.92%,严重程度评估准确率为100%。结论 该方法能够实现准确、高效且高度自动化的模型更新,降低主观性与计算复杂度,并为导管架结构的工程化智能FEMU提供了一条切实可行的路径。
Abstract
The work aims to propose a deep learning-based three-stage classification framework that enables intelligent, end-to-end FEMU for jacket structures to deal with the problems of low efficiency, limited automation, and subjective human decision-making in conventional updating approaches. The framework operationalized a "Qualify-Localize-Quantify (QLQ)" decision process to complete qualitative detection of whether updating was necessary, localization of the regions requiring updating, and quantification of the update severity. A shared feature extractor coupled with a dual-attention design learned informative representations directly from raw structural response data, allowing the system to automate the full updating workflow. Using a dataset of 83 421 load cases with train/validation/test splits of 7∶1.5∶1.5, the proposed method attained strong and consistent performance across all stages: 97.21% accuracy for update detection, 93.92% regional classification accuracy, and 100% accuracy for severity assessment. These results demonstrate that the approach delivers accurate, efficient, and highly automated updating, reducing subjectivity and computational overhead while offering a practical pathway to deploy intelligent FEMU in real-world jacket structures.
关键词
海洋工程 /
深度学习 /
有限元模型修正 /
多级分类 /
结构响应分析 /
智能修正 /
导管架结构
Key words
ocean engineering /
deep learning /
finite element model updating (FEMU) /
multi-stage classification /
structural response analysis /
intelligent updating /
jacket structures
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] FARRAR C R, WORDEN K.Structural Health Monitoring: A Machine Learning Perspective[M]. New York, USA: John Wiley & Sons, 2012.
[2] ÇATBAŞ F N, KIJEWSKI-CORREA T, AKTAN A E.Structural Identification of Constructed Systems: Approaches, Methods, and Technologies for Effective Practice of St-Id[M]. Reston, VA: American Society of Civil Engineers, 2013.
[3] 马红阳, 李辉, 贾子义. 一种抗冰结构在动态冰载荷作用下的缩尺模型设计方法[J]. 装备环境工程, 2025, 22(11): 78-84.
MA H Y, LI H, JIA Z Y.A Scaled Model Design Method for an Ice-Resistant Structure under Dynamic Ice Loads[J]. Equipment Environmental Engineering, 2025, 22(11): 78-84.
[4] CATBAS N, GOKCE H B, FRANGOPOL D M.Predictive Analysis by Incorporating Uncertainty through a Family of Models Calibrated with Structural Health- Monitoring Data[J]. Journal of Engineering Mechanics, 2013, 139(6): 712-723.
[5] SHAHIDI S G, PAKZAD S N.Effect of Measurement Noise and Excitation on Generalized Response Surface Model Updating[J]. Engineering Structures, 2014, 75: 51-62.
[6] IMREGUN M, VISSER W J.A Review of Model Updating Techniques[J]. The Shock and Vibration Digest, 1991, 23(1): 9-20.
[7] MOTTERSHEAD J E, FRISWELL M I.Model Updating in Structural Dynamics: A Survey[J]. Journal of Sound and Vibration, 1993, 167(2): 347-375.
[8] TONG M, LIANG Z, LEE G C.Correction Criteria of Finite Element Modeling in Structural Dynamics[J]. Journal of Engineering Mechanics, 1992, 118(4): 663-682.
[9] COLLINS J, KENNEDY B, HART G.Statistical Identification of Structures[C]//Proceedings of the 14th Structures, Structural Dynamics, and Materials Conference. Williamsburg: AIAA, 1973.
[10] CHEN J C, GARBA J A.Analytical Model Improvement Using Modal Test Results[J]. AIAA Journal, 1980, 18(6): 684-690.
[11] BARUCH M, BAR ITZHACK I Y. Optimal Weighted Orttiogonalization of Measured Modes[J]. AIAA Journal, 1978, 16(4): 346-351.
[12] BERMAN A, NAGY E J.Improvement of a Large Analytical Model Using Test Data[J]. AIAA Journal, 1983, 21(8): 1168-1173.
[13] BARUCH M.Methods of Reference Basis for Identification of Linear Dynamic Structures[J]. AIAA Journal, 1984, 22(4): 561-564.
[14] FRISWELL M I, MOTTERSHEAD J E.Finite Element Model Updating in Structural Dynamics[M]. Dordrecht: Springer Netherlands, 1995.
[15] SIPPLE J D, SANAYEI M.Finite Element Model Updating Using Frequency Response Functions and Numerical Sensitivities[J]. Structural Control and Health Monitoring, 2014, 21(5): 784-802.
[16] IMREGUN M, SANLITURK K Y, EWINS D J.Finite Element Model Updating Using Frequency Response Function Data II. Case Study on a Medium-Size Finite Element Model[J]. Mechanical Systems and Signal Processing, 1995, 9(2): 203-213.
[17] REZAIEE-PAJAND M, ENTEZAMI A, SARMADI H.A Sensitivity-Based Finite Element Model Updating Based on Unconstrained Optimization Problem and Regularized Solution Methods[J]. Structural Control and Health Monitoring, 2020, 27(5): e2481.
[18] MOTTERSHEAD J E, LINK M, FRISWELL M I.The Sensitivity Method in Finite Element Model Updating: A Tutorial[J]. Mechanical Systems and Signal Processing, 2011, 25(7): 2275-2296.
[19] GLADWELL G M L, AHMADIAN H. Generic Element Matrices Suitable for Finite Element Model Updating[J]. Mechanical Systems and Signal Processing, 1995, 9(6): 601-614.
[20] RATCLIFFE M J, LIEVEN N A J. A Generic element-Based Method for Joint Identification[J]. Mechanical Systems and Signal Processing, 2000, 14(1): 3-28.
[21] MOTTERSHEAD J E, FRISWELL M I, NG G H T, et al. Geometric Parameters for Finite Element Model Updating of Joints and Constraints[J]. Mechanical Systems and Signal Processing, 1996, 10(2): 171-182.
[22] AHMADIAN H, GLADWELL G M L, ISMAIL F. Parameter Selection Strategies in Finite Element Model Updating[J]. Journal of Vibration and Acoustics, 1997, 119(1): 37-45.
[23] JUNG H.Structural Dynamic Model Updating Using Eigensensitivity Analysis[D]. London: Imperial College London, 1992.
[24] O’CALLAHAN J C, AVITABILE P, RIEMER R. System Equivalent Reduction Expansion Process (SEREP)[C]//Proceedings of the 7th International Modal Analysis Conference. Las Vegas:[s. n.], 1989.
[25] CHU C H, TRETHEWEY M W.Rapid Structural Design Change Evaluation with an Experiment Based Fem[J]. Journal of Sound and Vibration, 1998, 211(3): 335-353.
[26] MARWALA T.Finite Element Model Updating Using Computational Intelligence Techniques: Applications to Structural Dynamics[M]. London: Springer London, 2010.
[27] LEVIN R I, LIEVEN N A J. Dynamic Finite Element Model Updating Using Simulated Annealing and Genetic Algorithms[J]. Mechanical Systems and Signal Processing, 1998, 12(1): 91-120.
[28] ZIMMERMAN D C, YAP K, HASSELMAN T.Evolutionary Approach for Model Refinement[J]. Mechanical Systems and Signal Processing, 1999, 13(4): 609-625.
[29] LI K Y, FANG J, SUN B, et al.Structural Dynamic Model Updating with Automatic Mode Identification Using Particle Swarm Optimization[J]. Applied Sciences, 2022, 12(18): 8958.
[30] LO IACONO F.Continuous Particle Swarm Optimization for Model Updating of Structures from Experimental Modal Analysis[J]. Materials Research Proceedings, 2023, 26: 467-472.
[31] GIRARDI M, PADOVANI C, PELLEGRINI D, et al.A Finite Element Model Updating Method Based on Global Optimization[J]. Mechanical Systems and Signal Processing, 2021, 152: 107372.
[32] 杨伟涛, 刘康泰, 邓鎏. 基于人工智能的弹药毁伤效能评估优化框架研究[J]. 装备环境工程, 2025, 22(9): 94-104.
YANG W T, LIU K T, DENG L.Optimization Framework for Ammunition Damage Effectiveness Evaluation Based on Artificial Intelligence[J]. Equipment Environmental Engineering, 2025, 22(9): 94-104.
[33] 杨涵, 高骏, 刘勇, 等. 激光雷达点云球面投影与相机融合的环境语义分割算法[J]. 装备环境工程, 2025, 22(7): 9-15.
YANG H, GAO J, LIU Y, et al.Environmental Semantic Segmentation Algorithm via LiDAR Point Cloud Spherical Projection and Camera Fusion[J]. Equipment Environmental Engineering, 2025, 22(7): 9-15.
[34] 丁雅杰, 王佐才, 辛宇, 等. 基于CNN-BiLSTM-Attention混合神经网络的结构非线性模型修正[J]. 工程力学, 2025, 41(5): 1-13.
DING Y J, WANG Z C, XIN Y, et al.Structural Nonlinear Model Updating Based on CNN-BiLSTM-Attention Hybrid Neural Network[J]. Engineering Mechanics, 2025, 41(5): 1-13.
[35] YUAN Z Q, KUANG X C, WANG Z C, et al.A Probabilistic Method for Structural Model Updating Using a Model-Data Hybrid Driven Technique[J]. Structures, 2025, 77: 109057.
[36] PARK H S, OH B K.CNN-Based Model Updating for Structures by Direct Use of Dynamic Structural Response Measurements[J]. Engineering Structures, 2024, 307: 117880.
PDF(4035 KB)
