融合密集卷积网络和注意力机制的拱桥损伤识别

PDF(3420 KB)

振动与冲击 ›› 2024, Vol. 43 ›› Issue (14) : 18-28.

论文

融合密集卷积网络和注意力机制的拱桥损伤识别

辛景舟1,2，刘倩茹2，唐启智1,2，李杰3，张洪1，周建庭1,2

作者信息 +

Damage identification of arch bridges based on dense convolutional networks and attention mechanisms

XIN Jingzhou1,2，LIU Qianru2，TANG Qizhi1,2，LI Jie3，ZHANG Hong1，ZHOU Jianting1,2

Author information +

文章历史 +

摘要

针对传统深度学习方法缺乏对网络特征的差异化利用、且损伤识别精度易受环境因素影响的问题，本文提出了一种融合密集卷积网络（DenseNet121）和注意力机制（CBAM）的拱桥损伤识别方法。首先，获取拱桥加速度响应数据，利用连续小波变换将其转换成时频图，形成拱桥损伤识别数据集；其次，将CBAM嵌入DenseNet121网络模型，加强断层特征的传播和特征的差异化利用，经训练得到拱桥损伤识别模型；然后，基于测试集评估损伤识别模型的性能，并引入t-SNE非线性降维技术对特征进行可视化分析；最后，通过数值案例验证了方法的可行性和鲁棒性，并应用于劲性骨架拱肋的损伤识别。结果表明，所提方法增强有用信息的权重，实现网络特征的差异化利用；与传统方法相比，该方法在单损伤和多损伤识别中准确率分别达到了91.67%和92.78%，准确率更高，且具有较强的鲁棒性和实用价值。

Abstract

Traditional deep learning methods lack the differentiated utilization of network features and the damage identification accuracy is easily affected by the environmental factors. To this end, this paper proposes an arch bridge damage identification method based on the dense convolutional network (DenseNet121) and the attention mechanism (CBAM). Firstly, the arch bridge acceleration data are obtained and converted into time-frequency maps using continuous wavelet transform to construct the damage identification dataset. Secondly, the CBAM is embedded into the DenseNet121 model to enhance the propagation of the fault features and the differential utilization of the features, and the arch bridge damage identification model is obtained after the training. Then, the performance of the damage identification model is evaluated based on the test set and the t-SNE nonlinear dimensionality reduction technique is introduced to visualize the features. Finally, the feasibility and robustness of the method are verified by numerical cases, and the proposed method is further applied to the damage identification of the stiff skeleton arch ribs. The results indicate that the proposed method can increase the weight of useful information and realize the differentiated utilization of network features. Compared with the traditional method, this method obtain a accuracy of 91.67% and 92.78% respectively in single damage identification and multi-damage identification, with higher identification accuracy, strong robustness, and practical value.

导出引用

辛景舟1, 2, 刘倩茹2, 唐启智1, 2, 李杰3, 张洪1, 周建庭1, 2. 融合密集卷积网络和注意力机制的拱桥损伤识别[J]. 振动与冲击, 2024, 43(14): 18-28

XIN Jingzhou1, 2, LIU Qianru2, TANG Qizhi1, 2, LI Jie3, ZHANG Hong1, ZHOU Jianting1, 2. Damage identification of arch bridges based on dense convolutional networks and attention mechanisms[J]. Journal of Vibration and Shock, 2024, 43(14): 18-28

参考文献

[1] Xin J Z, Jiang Y, Zhou J T, et al. Bridge deformation prediction based on SHM data using improved VMD and conditional KDE[J]. Engineering Structures, 2022, 261: 114285. [2] Xin J Z, Wang J Y, Liu Z W, et al. Experimental study on bearing capacity of corroded reinforced concrete arch considering material degradation[J]. Frontiers in Materials, 2022, 9: 847193. [3] 周建庭, 李晓庆, 辛景舟, 等. 基于Kalman-GARCH模型的结构损伤识别[J]. 振动与冲击, 2020, 39(6): 1-7+21. ZHOU Jian-ting, LI Xiao-qing, XIN Jing-zhou, et al. Structural damage identification based on a Kalman-GARCH model [J]. Journal of Vibration and Shock, 2020, 39(6): 1-7+21. [4] 唐启智, 辛景舟, 周建庭, 等. 基于时序分析的锈蚀RC拱损伤识别:仿真与验证[J]. 重庆交通大学学报(自然科学版), 2021, 40(2): 69-77. TANG Qi-zhi, XIN Jing-zhou, ZHOU Jian-ting, et al. Damage identification of corroded RC arch based on time series analysis: Simulation and verification [J]. Journal of Chongqing Jiaotong University (Natural Science Edition), 2021, 40(2): 69-77. [5] 缪炳荣, 杨树旺, 王名月, 等. 利用振动响应的多种结构损伤识别方法比较[J]. 振动工程学报, 2020, 33(4): 724-733. MIAO Bing-rong, YANG Shu-wang, WANG Ming-yue et al. Comparison of various structural damage identification methods using vibration response[J]. Journal of Vibration Engineering, 2020, 33(4): 724-733. [6] 姜绍飞, 胡春明. 基于模态指标与数据融合的钢管混凝土拱桥损伤识别[J]. 振动与冲击, 2009, 28(12): 91-95. JIANG Shao-fei, HU Chun-ming. Structural damage identification based on modal indices and data fusion for an arch bridge of concrete-filled steel tubes[J]. Journal of Vibration and Shock, 2009, 28(12): 91-95. [7] 谢开仲, 覃乐勤, 吕文高. 钢管混凝土拱桥拱肋模型修正方法研究[J]. 桥梁建设, 2010(3): 24-28. XIE Kai-zhong, QIN Le-qin, LU Wen-gao. Resrach of a model updating method for arch rib of CFST arch bridge[J]. Bridge Construction, 2010(3): 24-28. [8] Wang W W, Su M B, Wang C. Static deflection difference-based damage identification of hanger in arch bridges[J]. KSCE Journal of Civil Engineering, 2022, 26(12): 5096-5106. [9] Zhou S C, Jiang Y, Li X Q, et al. Vibration-based damage identification of reinforced concrete arch bridges using Kalman-ARMA-GARCH model[J]. International Journal of Robotics and Automation, 2021, 36(10): 206-0705. [10] 唐启智, 辛景舟, 周建庭，等. 基于AR-GP模型的结构损伤识别方法[J]. 振动与冲击, 2021，40(9): 102-109. TANG Qizhi, XIN Jingzhou, ZHOU Jianting, et al. Structural damage identification method based on ARGP model [J]. Journal of Vibration and Shock, 2021, 40(9): 102-109. [11] 刘春城, 刘佼. 基于支持向量机的大跨度拱桥损伤识别方法研究[J]. 振动与冲击, 2010, 29(7): 174-178+244. LIU Cheng-chun, LIU Jiao. Damage identification of a long-span arch bridge based on support vector machine[J]. Journal of Vibration and Shock, 2010, 29(7): 174-178+244. [12] Jayasundara N, Thambiratnam D P, Chan T H T., et al, Damage detection and quantification in deck type arch bridges using vibration based methods and artificial neural networks[J]. Engineering Failure Analysis, 2020, 109: 104265. [13] Conde B, Eguía P, Stavroulakis G E, et al. Parameter identification for damaged condition investigation on masonry arch bridges using a Bayesian approach[J]. Engineering Structures, 2018, 172: 275-284. [14] Yao J H, Zeng B, Zhou Z, et al. Damage identification analysis of cable-stayed arch-truss based on multi-node time-domain data fusion[J]. Latin American Journal of Solids and Structures, 2023, 20(1): e473. [15] Yu Y, Li J T, Li J C, et al. Automated damage diagnosis of concrete jack arch beam using optimized deep stacked autoencoders and multi-sensor fusion[J]. Developments in the Built Environment, 2023, 14: 100128. [16] Duan Y F, Chen Q Y, Zhang H M, et al. CNN-based damage identification method of tied-arch bridge using spatial-spectral information[J]. Smart Structures and Systems, 2019, 23(5): 507-520. [17] Almutairi M, Nikitas N, Abdeljaber O, et al. A methodological approach towards evaluating structural damage severity using 1D CNNs[J]. Structures, 2021, 34: 4435-4446. [18] 曲诚, 陈景龙, 常元洪, 等. 面向钢丝绳微弱损伤智能识别的多尺度注意力网络[J]. 西安交通大学学报, 2021, 55(07): 141-150. QU Cheng, CHEN Jing-long, CHANG Yuan-hong. Multi-Scale attention network for intelligent identification of weak damage on wire ropes[J]. Journal of Xi’an Jiaotong University, 2021, 55(07): 141-150. [19] 董绍江, 裴雪武, 吴文亮, 等. 改进抗干扰CNN的变负载滚动轴承损伤程度识别[J]. 振动、测试与诊断, 2021, 41(04): 715-722+831. DONG Shao-jiang, PEI Xue-wu, Wu Wen-liang. Damage degree identification of rolling bearings under variable load with improved anti⁃interference CNN[J]. Journal of Vibration，Measurement & Diagnosis, 2021, 41(04): 715-722+831. [20] 张健飞, 黄朝东, 王子凡. 基于多头自注意力机制和卷积神经网络的结构损伤识别研究[J]. 振动与冲击, 2022, 41(24): 60-71. ZHANG Jian-fei, HUANG Chao-dong, WANG Zi-fan. Research on structural damage identification based on multi⁃head self⁃attention and convolutional neural networks[J]. Journal of Vibration and Shock, 2022, 41(24): 60-71. [21] Fu L, Tang Q Z, Gao P, et al. Damage identification of long-span bridges using the hybrid of convolutional neural network and long short-term memory network[J]. Algorithms, 2021, 14(6): 180. [22] Tang Q Z, Xin J Z, Jiang Y, et al. Novel identification technique of moving loads using the random response power spectral density and deep transfer learning[J]. Measurement, 2022(195): 111120. [23] Scherer D, Müller A, Behnke S. Evaluation of pooling operations in convolutional architectures for object recognition[C]// International Conference on Artificial Neural Networks. Springer, Berlin, Heidelberg, 2010. [24] Wang R H, Li J, Chencho, et al. Densely connected convolutional networks for vibration based structural damage identification[J]. Engineering Structures, 2021, 245: 112871. [25] 张珂, 冯晓晗, 郭玉荣, 等. 图像分类的深度卷积神经网络模型综述[J]. 中国图象图形学报, 2021, 26(10): 2305-2325. ZHANG Ke, FENG Xiao-han, GUO Yurong, et al. Overview of deep convolutional neural networks for image classification[J]. Journal of Image and Graphics, 2021, 26(10): 2305-2325 [26] 姚齐水, 别帅帅, 余江鸿, 等. 一种结合改进Inception V2模块和CBAM的轴承故障诊断方法[J]. 振动工程学报, 2022, 35(4): 949-957. YAO Qi-shui, BIE Shuai-shuai, YU Jiang-hong, et al. A bearing fault diagnosis method combining improved Inception V2 module and CBAM[J]. Journal of Vibration Engineering, 2022, 35(4): 949-957.