Fixed-interval smoothing algorithm based on singular value decomposition

In this paper, a new fixed-interval smoothing algorithm based on singular value decomposition (SVD) is presented. The main idea of the new algorithm is to combine a forward-pass SVD-based square-root Kalman filter, developed recently by the authors, with a Rauch-Tung-Striebel backward-pass recursive smoother by using the SVD as a main computational tool. Similarly to the SVD-based square-root filter, the proposed smoother has good numerical stability and does not require covariance matrix inversion. It is formulated in a vector-matrix form, and thus is handy for implementation with parallel computers. A typical numerical example is used to demonstrate the performance of the new smoother.

[1]  E. G. Kogbetliantz.  Solution of linear equations by diagonalization of coefficients matrix , 1955 .

[2]  M. Hestenes.  Inversion of Matrices by Biorthogonalization and Related Results , 1958 .

[3]  C. Striebel,et al.  On the maximum likelihood estimates for linear dynamic systems , 1965 .

[4]  A. Bryson,et al.  DISCRETE SQUARE ROOT SMOOTHING , 1972 .

[5]  J. Meditch.  A survey of data smoothing for linear and nonlinear dynamic systems , 1973 .

[6]  Gerald Bierman,et al.  Sequential square root filtering and smoothing of discrete linear systems , 1974, Autom..

[7]  A. Laub,et al.  The singular value decomposition: Its computation and some applications , 1980 .

[8]  Gerald J. Bierman,et al.  A new computationally efficient fixed-interval, discrete-time smoother , 1981, 1981 20th IEEE Conference on Decision and Control including the Symposium on Adaptive Processes.

[9]  Keigo Watanabe.  A new forward-pass fixed-interval smoother using the U-D information matrix factorization , 1986, Autom..

[10]  Yaakov Oshman,et al.  Square root filtering via covariance and information eigenfactors , 1986, Autom..

[11]  S. G. Tzafestas,et al.  New computationally efficient formula for backward-pass fixed-interval smoother and its UD factorisation algorithm , 1989 .

[12]  Yaakov Oshman,et al.  Gain-free square root information filtering using the spectral decomposition , 1989 .

[13]  A. Sameh,et al.  An overview of parallel algorithms for the singular value and symmetric eigenvalue problems , 1989 .

[14]  S. R. McReynolds.  Covariance factorization algorithms for fixed-interval smoothing of linear discrete dynamic systems , 1990 .

[15]  Mi Lu,et al.  Parallel Computation of the Modified Extended Kalman Filter , 1991, ICPP.

[16]  Pierre Manneback,et al.  Kalman filter algorithm based on singular value decomposition , 1992, [1992] Proceedings of the 31st IEEE Conference on Decision and Control.

[17]  L. Wang,et al.  A singular value decomposition based Kalman filter algorithm , 1992, Proceedings of the 1992 International Conference on Industrial Electronics, Control, Instrumentation, and Automation.

[18]  A. Swindlehurst,et al.  Subspace-based signal analysis using singular value decomposition , 1993, Proc. IEEE.

[19]  Youmin Zhang,et al.  A SVD-based extended Kalman filter and applications to aircraft flight state and parameter estimation , 1994, Proceedings of 1994 American Control Conference - ACC '94.

[20]  A New Recursive Identification Method Based On Singular Value Decomposition , 1995 .

[21]  Thomas Kailath,et al.  Square-root RTS smoothing algorithms , 1995 .

[22]  T. Kailath,et al.  New square-root smoothing algorithms , 1996, IEEE Trans. Autom. Control..