A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Study of damage identification for bridges based on deep belief network
https://orcid.org/0000-0002-4671-1604
To solve the problem of poor anti-noise ability faced by traditional pattern recognition methods in damage identification field, a bridge damage identification method based on deep belief network was proposed. Taken the modal curvature difference as the damage index, three restricted Boltzmann machines were constructed for pre-training. Then, the Softmax classifier and neural network were used to identify the damage location and degree under the environmental cases of no noise, weak noise, and strong noise, respectively. Subsequently, the influence of incomplete measurement modal data on the method was studied. Finally, damage identification based on deep belief network was implemented to a continuous beam bridge and compared with that of the back propagation neural network. The results showed that the proposed method could be highly effective not only on damage location but also on degree identification. Compared with back propagation neural network, deep belief network method may possess better identification ability and stronger anti-noise ability. It also demonstrates good identification effect under the condition of incomplete measurement modal data.
Study of damage identification for bridges based on deep belief network
https://orcid.org/0000-0002-4671-1604
To solve the problem of poor anti-noise ability faced by traditional pattern recognition methods in damage identification field, a bridge damage identification method based on deep belief network was proposed. Taken the modal curvature difference as the damage index, three restricted Boltzmann machines were constructed for pre-training. Then, the Softmax classifier and neural network were used to identify the damage location and degree under the environmental cases of no noise, weak noise, and strong noise, respectively. Subsequently, the influence of incomplete measurement modal data on the method was studied. Finally, damage identification based on deep belief network was implemented to a continuous beam bridge and compared with that of the back propagation neural network. The results showed that the proposed method could be highly effective not only on damage location but also on degree identification. Compared with back propagation neural network, deep belief network method may possess better identification ability and stronger anti-noise ability. It also demonstrates good identification effect under the condition of incomplete measurement modal data.
Study of damage identification for bridges based on deep belief network
https://orcid.org/0000-0002-4671-1604
To solve the problem of poor anti-noise ability faced by traditional pattern recognition methods in damage identification field, a bridge damage identification method based on deep belief network was proposed. Taken the modal curvature difference as the damage index, three restricted Boltzmann machines were constructed for pre-training. Then, the Softmax classifier and neural network were used to identify the damage location and degree under the environmental cases of no noise, weak noise, and strong noise, respectively. Subsequently, the influence of incomplete measurement modal data on the method was studied. Finally, damage identification based on deep belief network was implemented to a continuous beam bridge and compared with that of the back propagation neural network. The results showed that the proposed method could be highly effective not only on damage location but also on degree identification. Compared with back propagation neural network, deep belief network method may possess better identification ability and stronger anti-noise ability. It also demonstrates good identification effect under the condition of incomplete measurement modal data.
Study of damage identification for bridges based on deep belief network
Guo, Qi (author) / Feng, Lei (author) / Zhang, Ruyi (author) / Yin, Haijun (author)
Advances in Structural Engineering ; 23 ; 1562-1572
2020-06-01
11 pages
Article (Journal)
Electronic Resource
English
Damage identification in bridges combining deep learning and computational mechanic
| BASE | 2022
Identification of Groundwater Contamination Sources Based on a Deep Belief Neural Network
| DOAJ | 2024
Deep Belief Network For Smoke Detection
| British Library Online Contents | 2017
Supervised Deep Learning with Finite Element simulations for damage identification in bridges
| BASE | 2022