Abstract
We consider the task of diagonalizing symmetric time varying matrices A(t). Based on the dynamic inversion technique developed by Getz and Marsden, a differential equation is proposed, whose solutions asymptotically track the diagonalizing transformation. In particular, one does not need to perfectly match the initial conditions, as the solutions converge exponentially towards the desired transformation. Thus, the proposed method is robust under perturbations.
Chapter PDF
Similar content being viewed by others
References
Bell, S.J.G., Nichols, N.K.: A numerical method for the computation of the analytic singular value decomposition, Numerical Analysis Report 2, 1994.
Braden, H.W.: The equations A′X +-X′A = B, SIAM Journal on Matrix Analysis and Applications, 20, (1998), 295–302.
Dehaene, J.: Continuous-Time Matrix Algorithms, Systolic Algorithms and Adaptive Neural Networks, Ph.D. Thesis KU Leuven, 1995.
Dieci, L., Eirola, T.: On smooth decompositions of matrices, SIAM Journal on Matrix Analysis and Applications, 20, (1999) 800–819.
Getz, N.H.: Dynamic Inversion of Nonlinear Maps, Ph.D. Thesis, Univ. of California at Berkeley, 1995.
Kato, T.: Perturbation Theory for Linear Operators, Springer-Verlag Berlin, 1966.
Wright, K.: Differential equations for the analytical singular value decomposition of a matrix, Numer. Math. 63:283 (1992).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Baumann, M., Helmke, U. (2002). Diagonalization of Time Varying Symmetric Matrices. In: Sloot, P.M.A., Hoekstra, A.G., Tan, C.J.K., Dongarra, J.J. (eds) Computational Science — ICCS 2002. ICCS 2002. Lecture Notes in Computer Science, vol 2331. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-47789-6_44
Download citation
DOI: https://doi.org/10.1007/3-540-47789-6_44
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-43594-5
Online ISBN: 978-3-540-47789-1
eBook Packages: Springer Book Archive