当前位置: X-MOL 学术KSCE J. Civ. Eng. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Structural Deformation Sensing Based on Distributed Optical Fiber Monitoring Technology and Neural Network
KSCE Journal of Civil Engineering ( IF 2.2 ) Pub Date : 2021-07-10 , DOI: 10.1007/s12205-021-1805-z
Gong-Yu Hou 1, 2 , Zi-Xiang Li 1 , Kai-Di Wang 1 , Jin-Xin Hu 1
Affiliation  

Structural deformation monitoring is vital to the safety of concrete structures. However, the distributed deformation of structures cannot be easily obtained using existing monitoring methods in civil engineering. To this end, this paper proposes a method to estimate the continuous deformation of concrete beams by utilizing the distributed optical fiber monitoring technology. In this method, optical fibers and a total station are used to obtain the strain and deformation distribution curves of a concrete beam, respectively. Subsequently, these curves are inputted to a back propagation network as training samples to learn their relationships. The results show that the deformation value of trained neural network is very close to that of the total station, with a maximum error of only 2.7% (0.3 mm). The linear regression analysis shows a goodness of fit R2 greater than 0.98, which confirms the reliability of the simulations results.



中文翻译:

基于分布式光纤监测技术和神经网络的结构变形传感

结构变形监测对混凝土结构的安全至关重要。然而,使用现有的土木工程监测方法无法轻易获得结构的分布变形。为此,本文提出了一种利用分布式光纤监测技术估算混凝土梁连续变形的方法。该方法利用光纤和全站仪分别获得混凝土梁的应变和变形分布曲线。随后,将这些曲线作为训练样本输入到反向传播网络中以了解它们之间的关系。结果表明,训练后的神经网络的变形值与全站仪非常接近,最大误差仅为2.7%(0.3 mm)。R 2大于0.98,证实了模拟结果的可靠性。

更新日期:2021-07-12
down
wechat
bug