Transactions of the Japan Society for Computational Engineering and Science
Online ISSN : 1347-8826
ISSN-L : 1344-9443
Characteristics of variable parametric projection filter using identification of lateral stiffness of 3-story frame models
Yoshihito IKEDARyuji ENDONobuyoshi TOSAKA
Author information
JOURNAL FREE ACCESS

2014 Volume 2014 Pages 20140009

Details
Abstract

Inverse analysis to identify every lateral stiffness of 3-story frame models was performed to study the characteristics of variable parametric projection filtering algorithm that is used as the inverse analysis procedure. Natural frequencies measured by the experimental modal analysis system were adopted as the observation data of the inverse problem. The variable parametric projection filter is including regularization parameter. As the regularization parameter is determined computationally by using natural frequency corresponding specific vibration mode, selection of vibration mode to determine the regularization parameter is important factor characterizing the variable parametric projection filter along with the initial value of iterative calculations. In this paper, characteristics of the variable parametric projection filter are made clear from viewpoint of the mode-dependent properties and initial value of iterative calculations through comparison with results of inverse analysis using Kalman filtering algorithm.

Content from these authors
© 2014 The Japan Society For Computational Engineering and Science
Previous article Next article
feedback
Top