Bridge Structural Damage Identification Technique Based on BPNN and Vehicle-bridge Interaction Analysis
During the use of bridges, the traditional method of detecting the bridge condition cannot be continuously monitored and maintained. To address this problem, the study proposed a damage identification method based on the interaction of back propagation neural network and vehicle-bridge interaction....
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Electronic Journals for Science and Engineering - International
2025-07-01
|
| Series: | Electronic Journal of Structural Engineering |
| Subjects: | |
| Online Access: | http://10.0.0.97/EJSE/article/view/774 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | During the use of bridges, the traditional method of detecting the bridge condition cannot be continuously monitored and maintained. To address this problem, the study proposed a damage identification method based on the interaction of back propagation neural network and vehicle-bridge interaction. The method analyzed the car response when the car passes over the bridge through back propagation neural network combined with coupled vibration of the vehicle-bridge, and carried out the inference of the response of the contact point of the car tire and the bridge. To create the damaged structural response model of the simulated bridge, the stiffness of the bridge contact point unit was then decreased. The response of the contact point of the bridge deck and tires was then used as the input of the back propagation neural network to compute the coupled vibration equations of the vehicle and bridge, and create the data set of the vehicle and bridge’s response. It can also locate the damaged bridge structure appropriately, and assess the extent of damage. The results demonstrated that the average accuracy of back propagation neural network in locating the damaged bridge structure was about 90%, the average accuracy of locating the damaged structure was kept at 85% under circumstances where varying noise levels are present. The maximum accuracy of assessing the degree of damage to the damaged structure was 98.54%, which is around 10% greater than the deep belief network and support vector machine’s performance in identifying the damage to the bridge structure. Taken together, the proposed bridge structure damage identification method can achieve high localization accuracy as well as quantitative accuracy.
|
|---|---|
| ISSN: | 1443-9255 |