首页 | 本学科首页   官方微博 | 高级检索  
     检索      

苏通大桥结构健康状态评估技术研究与应用(1):拉索损伤识别
引用本文:杨杰,李爱群,丁幼亮,姚蓓,杨军,张维苏.苏通大桥结构健康状态评估技术研究与应用(1):拉索损伤识别[J].地震学刊,2010(3):325-329.
作者姓名:杨杰  李爱群  丁幼亮  姚蓓  杨军  张维苏
作者单位:[1]东南大学混凝土与预应力混凝土结构教育部重点实验室,南京210096 [2]南京航空航天大学土木工程系,南京210016 [3]江苏省苏通大桥建设指挥部,南京210006
基金项目:国家杰出青年科学基金项目(50725828)、国家“十一五”科技支撑计划项目(2006BAJ03B05)、国家航空科学基金项目(2008ZD52040)资助
摘    要:采用改进的RBF神经网络建立了苏通大桥拉索损伤识别方法。2个不同阶固有频率之比是仅与损伤位置有关的结构振动参数,据此定义了用于损伤定位的损伤特征指标,并用其来训练神经网络;提出了基于R+^2准则与Jackknife校验的改进RBF算法,以有效地控制RBF网络的过拟合现象。算例结果表明,所提出的方法可以较好地对苏通大桥斜拉索进行损伤识别。

关 键 词:斜拉桥  结构损伤识别  神经网络  斜拉索

Technology and Application of Structural Health Condition Assessment for Sutong Bridge (1): Damage Identification of Stay Cable
YANG Jie,LI Ai-qun,DING You-liang,YAO Bei,YANG Jun,ZHANG Wei-su.Technology and Application of Structural Health Condition Assessment for Sutong Bridge (1): Damage Identification of Stay Cable[J].Journal of Seismology,2010(3):325-329.
Authors:YANG Jie  LI Ai-qun  DING You-liang  YAO Bei  YANG Jun  ZHANG Wei-su
Institution:1. Key Laboratory of Concrete and Prestressed Concrete Structures of Ministry of Education, Southeast University, Nanjing 210096, China; 2. Naniing University of Aeronautics and Astronautics, Nanjing 210016, China; 3. Construction Headquarter of Jiangsu Sutong Bridge, Nanjing 210006, China)
Abstract:In this paper the damage identification method for stay cable of Sutong Bridge is established using the modified RBF neural networks. The ratio of two different frequencies of the main girder is a vibration character which varies only with the damage location. It is selected as damage analysis index to train the neural networks. And in order to control the over-fitting of RBF neural networks, a new modified algorithm is presented based on R~_ rule and Jackknife validation. The analysis results show that the presented method can effectively locate the cable damage of Sutong Bridge.
Keywords:cable-stayed bridge  structural damage identification  neural network  stay cable
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号