基于改进广义卡尔曼滤波的结构损伤识别方法

英文篇名：Structural damage identification methods based on improved generalized Kalman filter
中文刊名：地震工程与工程振动
英文刊名：Journal of Earthquake Engineering and Engineering Vibration
作者：杜飞平 ; 谭永华 ; 陈建华
英文作者：DU Feiping1 ; TAN Yonghua2 ; CHEN Jianhua1 (1.Xi'an Aerospace Propulsion Institute ; Xi'an 710100 ; China ; 2.Academy of Aerospace Propulsion Technology ; Xi'an 710100 ; China)
中文关键词：广义卡尔曼滤波 ; 衰减记忆滤波 ; 自适应追踪滤波 ; 结构损伤
英文关键词：generalized Kalman filter ; fading Kalman filter ; adaptive tracking filter ; structural damage
出版日期：2010-08-15
机构：西安航天动力研究所;航天推进技术研究院;
年：2010
期：04
出版单位：地震工程与工程振动

摘要

针对传统的广义卡尔曼滤波算法不能有效地追踪结构刚度的变化情况,本文以传统卡尔曼滤波理论为基础,得到了基于衰减记忆的广义卡尔曼滤波算法公式,利用该算法对所得到的地震响应信号进行分析,提取结构的特性,辨识结构的参数,并从中判断结构损伤发生的时刻、位置及程度,改善了广义卡尔曼滤波的效果。衰减记忆的广义卡尔曼滤波算法只能够判断出结构参数变化的时间并且容易出现振荡,因此采用了一种新的自适应追踪技术,用一个自适应因子矩阵代替了原有的遗忘因子,这种技术可以有效追踪结构参数变化的时间、位置和大小,从而能够在线识别出结构的损伤。
Considering that the traditional generalized Kalman filter algorithm can not effectively track the changes in stiffness,based on the traditional Kalman filter,this paper works out the formulas of the fading Kalman filter.This paper uses the algorithm to analyze the responses of the earthquake signals to extract the characteristics of the structure,to identify the structural parameters,and determine the time,location and extent of the structural damage.The algorithm improves the generalized Kalman filtering effect.But fading Kalman filter algorithm can only determine the time when the structural parameters change,and may be prone to oscillation.This paper uses a new adaptive tracking technology,with an adaptive factor matrix instead of the original forgetting factor.The technology can effectively track the changes in structural parameters of time,location and extent,and thus applies to on-line identification.

引文

[1]孙鸿敏,李宏男.土木工程结构健康监测研究进展[J].防灾减灾工程学报,2003,23(3),92-98.
    [2]Houser G W,Bergman L A,Caughey TK.Structural control:past,present,and future[J].Journal of Engineering Mechanics,ASCE,1997,123(9):897-971.
    [3]黄道琼,张继桐,何洪庆.四机并联发动机低频动态特性分析[J].火箭推进,2004,30(4):27-31.
    [4]张惠军.液体火箭发动机故障检测与诊断技术综述[J].火箭推进,2004,30(5):40-45.
    [5]徐洪钟,吴中如.基于小波分析的大坝观测数据异常值检测[J].水电能源科学,2002,20(4):20-21.
    [6]Hoshiya M,Saito E.Structural identification by extended Kalman filter[J].Journal of Engineering Mechanics(ASCE),1984,110(12):1757-1771.
    [7]龙正平,刘卫东.一种自适应指数加权衰减记忆滤波算法[J].指挥控制与仿真,2006,28(6):41-45.
    [8]方青,梅晓春,张育平.用于机动目标跟踪的Kalman滤波器的设计[J].雷达科学与技术,2006,4(1):50-55.
    [9]Yang J N,Lin S.Identification of parametric variations of structures based on least squares estimation and adaptive tracking technique[J].Journalof Engineering Mechanics,ASCE,2005,131(3),290-298.
    [10]Yang J N,Pan S,Huang H.An adaptive extended Kalman filter for structural damage identificatons II:unknown inputs[J].Structural Controland Health Monitoring,2005,14(3):497-521.