基于径向基函数神经网络的斜拉桥损伤识别

为寻求桥梁结构自动损伤识别的方法，利用径向基函数（RBF）神经网络对某斜拉桥进行了损伤识别研究。分别采用了频率、振型模态、曲率模态3种指标作为网络的输入参数，考虑1根斜拉索损伤、2根斜拉索损伤及3根斜拉索损伤的三类工况，提出了损伤位置识别判断准则及识别效果评价指标。研究表明，径向基函数神经网络对斜拉桥的损伤位置和损伤程度能进行有效识别，构造样本和选择损伤指标是今后的研究方向。

Damage Detection of Cable-stayed Bridge Based on RBF Neural Networks

In order to find the methods of automatic damage detection for bridges, the damage detection of cable-stayed bridge is studied by using RBF neural network. In this modal, the frequencies, mode shapes, curvature mode are used as RBF neural networks import vector respectively. The research is undertaken in three instances, that are one cable of the bridge is damaged, two cables and three cables are damaged. The damage localization criteria and evaluation indices of identification effect are presented. The result indicates that RBF neural networks can detect not only the damage position but also the damage degree effectively. It is pointed out that the construction of samples and the choice of damage indices are the key technique in the model. 6 tabs, 3 figs, 6 refs.