地震动反演及结构参数识别的EKF算法  

INVERSION OF GROUND MOTION AND IDENTIFICATION OF STRUCTURAL PARAMETERS BY EKF

作　　者：王晓燕[1] 黄维平[1] 李华军[1] 

机构地区：[1]中国海洋大学海岸与海洋工程研究所,青岛266071

出　　处：《工程力学》2005年第4期20-23,111,共5页Engineering Mechanics

基　　金：国家自然科学基金资助项目(50179014);"十五"863计划资助项目(2001AA602023)

摘　　要：讨论了信息不完备的情况下剪切型结构的参数识别问题,利用不完备的测量信息,在系统载荷未知的条件下,采用最小二乘方法反演得到地震动响应时程,用扩展卡尔曼滤波(EKF)算法识别结构的动态物理参数,解决了扩展卡尔曼滤波算法需要输入的问题,取得了满意的参数识别结果。系统仿真算例表明,在引入20%的白噪声条件下,识别结果仍具有较高的计算精度,说明EKF算法用于剪切型结构参数识别时有较强的鲁棒性。Parameter identification of shear structure with incomplete information is discussed in this paper. The time history of ground acceleration has been inversed based on least square method under incomplete response data. Structural parameters are identified by Extended Kalman Filter （EKF）. The input information problem of EKF is resolved and good results are obtained. Numerical example shows that the identified results with small error are obtained under up to 20% of white noise. The method is shown to be robust and error-tolerant in identification of shear structures.

关 键 词：结构工程 参数识别 扩展卡尔曼滤波 载荷识别 地震反演 

分 类 号：TU312[建筑科学—结构工程]