Abstract
Alternating projection algorithms for matrix approximation are applied for structural model updating. Desired matrix properties such as sparsity, definiteness, and satisfaction of eigenconstraints are imposed as side constraints for a minimisation problem formulated to produce an updated matrix model which better matches measured data. Included are formulations of the update problem and discussion of the determination of solutions, convergence of the alternating projections approach as seen through geometric interpretations, and experimental verification. These approaches are verified through an application to locate damage in a laboratory truss structure using stiffness matrix approximation with experimental vibration measurements.
Original language | English |
---|---|
Pages (from-to) | 187-201 |
Number of pages | 15 |
Journal | Mechanical Systems and Signal Processing |
Volume | 12 |
Issue number | 1 |
DOIs | |
State | Published - Jan 1998 |
Bibliographical note
Funding Information:This work was supported in part by NASA grant NAG!0!0135[ The author appreciates conversations with Christopher A[ Beattie\ Nicholas J[ Higham\ and Buddy Glunt that contributed considerably to this work[
ASJC Scopus subject areas
- Control and Systems Engineering
- Signal Processing
- Civil and Structural Engineering
- Aerospace Engineering
- Mechanical Engineering
- Computer Science Applications