拡張可変的パラメトリック射影フィルタを援用した構造システム同定

Abstract

System identification was performed for five-story frame models to estimate lateral stiffness of every story at once in engineering field of inverse problem. In this study filtering algorithm based on variable parametric projection filter (VPPF) was used for the structural system identification adopted natural frequencies as observations. Identification results based on VPPF were compared with the results of Kalman filtering algorithm, the effectiveness of algorithm are made clear through the results on inverse analysis of various case from points of view on institution procedures of initial values and regularization parameter.