A Study of Sparse Matrix Formulation on the Experimental Spatial Matrix Identification Method

This paper presents a new development in the experimental method for identifying the set of spatial matrices. That is, to realize the identification on the spatial matrices in sparse matrix form in accordance with consideration of structural connections between measurement points. In the previous method that the authors proposed, spatial matrices are usually identified in full matrix form due to the difficulty of realizing sparse matrix. In this paper, the difficulty is broken through using the simplex method of nonlinear optimization. A simple beam structure is identified in sparse matrix form by the newly developed method and in full matrix form by the previous one. Some static and dynamic prediction analyses are then carried out to compare both methods. That is, to verify the validity and the usefulness of the spatial matrices identified by both methods. The results show some advantages of the new method.