Structural Damage Detection with Automatic Feature-Extraction through Deep Learning

被引:497
作者
Lin, Yi-zhou [1 ,2 ]
Nie, Zhen-hua [1 ,2 ]
Ma, Hong-wei [3 ,4 ]
机构
[1] Jinan Univ, Sch Mech & Construct Engn, Guangzhou, Guangdong, Peoples R China
[2] Minist Educ, Key Lab Disaster Forecast & Control Engn, Guangzhou, Guangdong, Peoples R China
[3] Dongguan Univ Technol, Dongguan, Peoples R China
[4] Qinghai Univ, Dept Civil Engn, Xining, Qinghai, Peoples R China
基金
中国国家自然科学基金;
关键词
NEURAL NETWORKS; IDENTIFICATION; BRIDGES;
D O I
10.1111/mice.12313
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Structural damage detection is still a challenging problem owing to the difficulty of extracting damage-sensitive and noise-robust features from structure response. This article presents a novel damage detection approach to automatically extract features from low-level sensor data through deep learning. A deep convolutional neural network is designed to learn features and identify damage locations, leading to an excellent localization accuracy on both noise-free and noisy data set, in contrast to another detector using wavelet packet component energy as the input feature. Visualization of the features learned by hidden layers in the network is implemented to get a physical insight into how the network works. It is found the learned features evolve with the depth from rough filters to the concept of vibration mode, implying the good performance results from its ability to learn essential characteristics behind the data.
引用
收藏
页码:1025 / 1046
页数:22
相关论文
共 52 条
[51]   Hilbert-Huang based approach for structural damage detection [J].
Yang, JN ;
Lei, Y ;
Lin, S ;
Huang, N .
JOURNAL OF ENGINEERING MECHANICS, 2004, 130 (01) :85-95
[52]  
Zhou Y. T., 1988, IEEE International Conference on Neural Networks (IEEE Cat. No.88CH2632-8), P71, DOI 10.1109/ICNN.1988.23914