Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network
This paper presents a novel approach to substantially improve the detection accuracy of structural damage via a one-dimensional convolutional neural network (1-D CNN) and a decision-level fusion strategy. As structural damage usually induces changes in the dynamic responses of a structure, a CNN can...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/12/3950 |
| _version_ | 1797530925643333632 |
|---|---|
| author | Shuai Teng Gongfa Chen Zongchao Liu Li Cheng Xiaoli Sun |
| author_facet | Shuai Teng Gongfa Chen Zongchao Liu Li Cheng Xiaoli Sun |
| author_sort | Shuai Teng |
| collection | DOAJ |
| description | This paper presents a novel approach to substantially improve the detection accuracy of structural damage via a one-dimensional convolutional neural network (1-D CNN) and a decision-level fusion strategy. As structural damage usually induces changes in the dynamic responses of a structure, a CNN can effectively extract structural damage information from the vibration signals and classify them into the corresponding damage categories. However, it is difficult to build a large-scale sensor system in practical engineering; the collected vibration signals are usually non-synchronous and contain incomplete structure information, resulting in some evident errors in the decision stage of the CNN. In this study, the acceleration signals of multiple acquisition points were obtained, and the signals of each acquisition point were used to train a 1-D CNN, and their performances were evaluated by using the corresponding testing samples. Subsequently, the prediction results of all CNNs were fused (decision-level fusion) to obtain the integrated detection results. This method was validated using both numerical and experimental models and compared with a control experiment (data-level fusion) in which all the acceleration signals were used to train a CNN. The results confirmed that: by fusing the prediction results of multiple CNN models, the detection accuracy was significantly improved; for the numerical and experimental models, the detection accuracy was 10% and 16–30%, respectively, higher than that of the control experiment. It was demonstrated that: training a CNN using the acceleration signals of each acquisition point and making its own decision (the CNN output) and then fusing these decisions could effectively improve the accuracy of damage detection of the CNN. |
| first_indexed | 2024-03-10T10:36:58Z |
| format | Article |
| id | doaj.art-3f17e19e8a0a4f33813f893deb7b8ba3 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T10:36:58Z |
| publishDate | 2021-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-3f17e19e8a0a4f33813f893deb7b8ba32023-11-21T23:14:00ZengMDPI AGSensors1424-82202021-06-012112395010.3390/s21123950Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural NetworkShuai Teng0Gongfa Chen1Zongchao Liu2Li Cheng3Xiaoli Sun4School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou 510006, ChinaSchool of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou 510006, ChinaSchool of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou 510006, ChinaDepartment of Mechanical Engineering, The Hong Kong Polytechnic University, Hong Kong 999077, ChinaSchool of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou 510006, ChinaThis paper presents a novel approach to substantially improve the detection accuracy of structural damage via a one-dimensional convolutional neural network (1-D CNN) and a decision-level fusion strategy. As structural damage usually induces changes in the dynamic responses of a structure, a CNN can effectively extract structural damage information from the vibration signals and classify them into the corresponding damage categories. However, it is difficult to build a large-scale sensor system in practical engineering; the collected vibration signals are usually non-synchronous and contain incomplete structure information, resulting in some evident errors in the decision stage of the CNN. In this study, the acceleration signals of multiple acquisition points were obtained, and the signals of each acquisition point were used to train a 1-D CNN, and their performances were evaluated by using the corresponding testing samples. Subsequently, the prediction results of all CNNs were fused (decision-level fusion) to obtain the integrated detection results. This method was validated using both numerical and experimental models and compared with a control experiment (data-level fusion) in which all the acceleration signals were used to train a CNN. The results confirmed that: by fusing the prediction results of multiple CNN models, the detection accuracy was significantly improved; for the numerical and experimental models, the detection accuracy was 10% and 16–30%, respectively, higher than that of the control experiment. It was demonstrated that: training a CNN using the acceleration signals of each acquisition point and making its own decision (the CNN output) and then fusing these decisions could effectively improve the accuracy of damage detection of the CNN.https://www.mdpi.com/1424-8220/21/12/3950structural damage detectiondecision-level fusion1-D convolutional neural networkvibration experimentsacceleration signalsbridge model |
| spellingShingle | Shuai Teng Gongfa Chen Zongchao Liu Li Cheng Xiaoli Sun Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network Sensors structural damage detection decision-level fusion 1-D convolutional neural network vibration experiments acceleration signals bridge model |
| title | Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network |
| title_full | Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network |
| title_fullStr | Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network |
| title_full_unstemmed | Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network |
| title_short | Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network |
| title_sort | multi sensor and decision level fusion based structural damage detection using a one dimensional convolutional neural network |
| topic | structural damage detection decision-level fusion 1-D convolutional neural network vibration experiments acceleration signals bridge model |
| url | https://www.mdpi.com/1424-8220/21/12/3950 |
| work_keys_str_mv | AT shuaiteng multisensoranddecisionlevelfusionbasedstructuraldamagedetectionusingaonedimensionalconvolutionalneuralnetwork AT gongfachen multisensoranddecisionlevelfusionbasedstructuraldamagedetectionusingaonedimensionalconvolutionalneuralnetwork AT zongchaoliu multisensoranddecisionlevelfusionbasedstructuraldamagedetectionusingaonedimensionalconvolutionalneuralnetwork AT licheng multisensoranddecisionlevelfusionbasedstructuraldamagedetectionusingaonedimensionalconvolutionalneuralnetwork AT xiaolisun multisensoranddecisionlevelfusionbasedstructuraldamagedetectionusingaonedimensionalconvolutionalneuralnetwork |