摘要
为提高对悬索桥损伤结构的测量精度,提出将吊索张力指标和神经网络技术相结合对悬索桥结构损伤识别的方法.基于高精度三维有限元模型,模拟7种可能的损伤情况的定位.采用BP神经网络,以不同损伤程度下吊索张力指标作为神经网络的训练与测试输入,由神经网络的输出来指示损伤位置及程度.该方法的突出优点是只利用少量吊索的局部模态的基频,就可获得较好的识别结果.而对少量吊索局部模态的基频测量要比其他面向损伤检测的测量容易得多.因此,该方法具有重要的实用价值.
A damage locating method for suspension bridge combining cable tension index and neural network technique was developed to improve the measuring accuracy for damaged structure. By using BP network, damage localization for 7 potential damage cases were simulated based on a high-precision three-dimensional finite element model. Taking cable tension indices as inputs of neural network for both training and testing, damage locations were indicated by the outputs of the network. The outstanding advantage of the method is that a good result can be obtained by measuring only fundamental natural frequencies of a few suspended cables instead of some other damage-oriented measurement which is more difficult than the former. The result shows that the method has great practical value.
出处
《大连海事大学学报》
EI
CAS
CSCD
北大核心
2008年第4期132-134,139,共4页
Journal of Dalian Maritime University
关键词
斜拉桥
损伤位置识别
斜拉索张力指标
神经网络
suspension bridge
damage location identification
cable tension index
neural network
