A NON-PARAMETRIC IDENTIFICATION METHODOLOGY FOR LARGE-SCALE NONLINEAR DYNAMIC SYSTEMS WITH UNCERTAIN MEASUREMENTS
In the field of structural dynamics,the data recorded by measurement sensors are often polluted by measurement uncertainty,commonly referred to as noise. The effect of the measurement uncertainty is frequently oversimplified to be simple additive Gaussian noise in many studies. In addition,for civil structures,such as large-scale nonlinear viscous dampers,stochastic quantification,which requires the collection of a large number of data sets,is often infeasible due to high costs associated with repeated physical testing even in laboratories equipped with large loading frames. A methodology is proposed for change detection of a dynamic model in the presence of uncertain measurements with limited physical data. It was demonstrated that reliable change detection is possible with the Restoring Force Method,a nonparametric system identification technique,because: (1) small changes of a nonlinear system are detectable; (2) physical interpretations are possible; and (3) the uncertainty of detected changes are quantifiable without a priori system characteristic knowledge.
Hae-Bum Yun Aaron Rank Sami F.Masri Gianmario Benzoni
Department of Civil, Environmental and Construction Engineering, University of Central Florida,Orlan Department of Civil and Environmental Engineering,University of Southern California,Los Angeles,CA 9 Department of Structural Engineering,University of California,San Diego,CA 92093,USA
国际会议
厦门
英文
1358-1364
2011-10-28（万方平台首次上网日期，不代表论文的发表时间）