Research on Two-Stage Damage Identification of Steel Frame Based on CNN and CMCM

WAN Neng; HUANG Minshui; ZHU Hongping

doi:10.3724/j.gyjzG23072612

Volume 54 Issue 1

Jan. 2024

Turn off MathJax

Article Contents

Article Navigation > INDUSTRIAL CONSTRUCTION > 2024 > 54(1): 123-129

WAN Neng, HUANG Minshui, ZHU Hongping. Research on Two-Stage Damage Identification of Steel Frame Based on CNN and CMCM[J]. INDUSTRIAL CONSTRUCTION, 2024, 54(1): 123-129. doi: 10.3724/j.gyjzG23072612

Citation:

WAN Neng, HUANG Minshui, ZHU Hongping. Research on Two-Stage Damage Identification of Steel Frame Based on CNN and CMCM[J]. INDUSTRIAL CONSTRUCTION, 2024, 54(1): 123-129. doi: 10.3724/j.gyjzG23072612

WAN Neng, HUANG Minshui, ZHU Hongping. Research on Two-Stage Damage Identification of Steel Frame Based on CNN and CMCM[J]. INDUSTRIAL CONSTRUCTION, 2024, 54(1): 123-129. doi: 10.3724/j.gyjzG23072612

Citation:

PDF( 1579 KB)

Research on Two-Stage Damage Identification of Steel Frame Based on CNN and CMCM

doi: 10.3724/j.gyjzG23072612

1. School of Civil Engineering and Architecture Wuhan Institute of Technology, Wuhan 430073, China;
2. School of Civil and Hydraulic Engineering Huazhong University of Science and Technology, Wuhan 430074, China

Received Date: 2023-07-26
Available Online: 2024-02-27

Abstract

Abstract

In the field of structural damage identification, the Cross Model Cross Mode (CMCM) Method is constrained by the rank deficiency of the coefficient matrix when solving noise-inclusive damage identification problems, leading to a decrease in the accuracy of damage identification results. To address this issue, this study proposes a two-stage damage identification method aimed at reducing the redundant equations in the CMCM method’s coefficient matrix, thereby enhancing the computational performance and noise robustness of the CMCM method. In the first stage, a Convolutional Neural Network (CNN) is employed for structural damage localization to eliminate the redundant equations in the coefficient matrix of the CMCM method. Subsequently, in the second stage, the reduced CMCM coefficient matrix equation is solved to obtain more accurate damage identification results. The effectiveness of the proposed two-stage damage identification method is validated through numerical and experimental studies. Compared with the traditional CMCM method, the method proposed in this paper significantly improves the accuracy of damage identification in solving noise-inclusive damage identification problems, demonstrating its superior performance.
- Structural damage identification,
- CMCM,
- CNN,
- a two-stage method

FullText(HTML)

References(19)

References

[1]	周奎, 王琦, 刘卫东, 等. 土木工程结构健康监测的研究进展综述[J]. 工业建筑, 2009, 39(3): 96-102.
[2]	M GATTI, Structural health monitoring of an operational bridge: A case study [J], Engineering Structures 195, 200209(2019).
[3]	温利明, 黄奕辉. 神经网络用于结构损伤识别的几个关键问题研究[J]. 工业建筑, 2002, 32(8): 39-40 , 49.
[4]	顾箭峰, 向春燕, 陶甫先等. 基于深度学习和IHPO的桥梁结构模型修正方法[J]. 广西大学学报(自然科学版), 2022, 47(05): 1147-1159.
[5]	翁顺, 朱宏平. 基于有限元模型修正的土木结构损伤识别方法[J]. 工程力学, 2021, 38(3): 1-16.
[6]	雷勇志, 黄民水, 顾箭峰, 等. 环境温度影响下基于支持向量机与强化飞蛾扑火优化算法的结构稀疏损伤识别[J]. 计算力学学报, 2022, 39(2): 170-177.
[7]	S L J HU, H LI, and S WANG, Cross-model cross-mode method for model updating [J], Mechanical Systems and Signal Processing 21, 416901703(2007).
[8]	H LI, J WANG, and S L JAMES HU, Using incomplete modal data for damage detection in offshore jacket structures [J], Ocean Engineering 35, 17-1817931799(2008).
[9]	贾辉. 基于改进交叉模型交叉模态法的损伤识别研究[D]. 大连: 大连理工大学, 2011.
[10]	占超, 李东升, 任亮, 等. 基于改进交叉模型交叉模态法的局部损伤识别方法[J]. 振动与冲击, 2015, 34(7): 127-133.
[11]	王炎, 陈辉, 黄斌, 等. 利用改进交叉模型交叉模态的随机模型修正方法[J]. 振动工程学报, 2023, 36(2): 498-506.
[12]	CONG S, HU S L J, LI H J. FRF-based pole-zero method for finite element model updating [J/OL]. Mechanical Systems and Signal Processing, 2022, 177[2022-05-05]. https://doi.org/10.1016/j.ymssp.2022.109206.
[13]	XU M, WANG S, JIANG Y. Structural damage identification by a cross modal energy sensitivity based mode subset selection strategy [J/OL]. Marine Structures, 2021, 77[2021-02-24]. https://doi.org/10.1016/j.marstruc.2021.102968.
[14]	K LIU, R J YAN, and C GUEDES SOARES, An improved model updating technique based on modal data [J], Ocean Engineering 154, 277287(2018).
[15]	LE C Y, BENGIO Y, HINTON G. Deep learning [J]. Nature, 2015, 521(7553): 436-444.
[16]	X Q ZHOU, Y XIA, and S WENG, L1, Structural Health Monitoring 14, 6571582(2015).
[17]	TIKHONOV A N, ARSENIN V Y. Solutions of ill-posed problems [M]. Preston: Winston Press, 1977.
[18]	E A JOHNSON, H F LAM, and L S KATAFYGIOTIS, et al.Phase I IASC-ASCE structural health monitoring benchmark problem using simulated data [J], Journal of engineering mechanics 130, 1315(2004).
[19]	OMENZETTER P, LAUTOUR O R D. Detection of Seismic Damage in Buildings Using Structural Responses: UNI/535[R]. Wellington: Earthquake Commission of New Zealand, 2008.