J191023 System Identification with Grouped Elements of Mass and Stiffness Matrices
Junya YAMAGUCHI, Masayoshi Misawa
Abstract
Open-access reader
Junya YAMAGUCHI, Masayoshi Misawa
Abstract
Open-access reader
This paper describes system identification to reduce modeling errors in analytical mass and stiffness matrices of structures. The mass and stiffness matrices including modeling errors can not provide the dynamic characteristics of real structures. Accordingly, we use an optimization method with constraints to identify the mass and stiffness matrices. In order to keep mass and stiffness distributions, element mass and stiffness matrices are divided into some groups and grouped matrices are modified every each group. In this method, different groups affect the precision of identification. Therefore, how to set groups is significantly important to accurately obtain the identified mass and stiffness matrices. In this paper, new groups are used for system identification. The mass and stiffness matrices are grouped considering modes of the structure. Numerical example shows that the use of new groups enables to accurately identify the mass and stiffness matrices and the identified frequencies agree with the simulated test frequencies.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper describes system identification to reduce modeling errors in analytical mass and stiffness matrices of structures. The mass and stiffness matrices including modeling errors can not provide the dynamic characteristics of real structures. Accordingly, we use an optimization method with constraints to identify the mass and stiffness matrices. In order to keep mass and stiffness distributions, element mass and stiffness matrices are divided into some groups and grouped matrices are modified every each group. In this method, different groups affect the precision of identification. Therefore, how to set groups is significantly important to accurately obtain the identified mass and stiffness matrices. In this paper, new groups are used for system identification. The mass and stiffness matrices are grouped considering modes of the structure. Numerical example shows that the use of new groups enables to accurately identify the mass and stiffness matrices and the identified frequencies agree with the simulated test frequencies.
Key concepts: Stiffness, Direct stiffness method, Identification (biology), Set (abstract data type), Structural engineering, Stiffness matrix, Mathematics, Computer science