Tensor network Kalman filter for LTI systems

Conference Paper (2019)
Author(s)

Daniel Gedon (Student TU Delft)

P.J. Piscaer (TU Delft - Team Raf Van de Plas)

K. Batselier (TU Delft - Team Jan-Willem van Wingerden)

S Smith (TU Delft - Team Raf Van de Plas)

M Verhaegen (TU Delft - Team Raf Van de Plas)

Research Group
Team Raf Van de Plas
Copyright
© 2019 Daniel Gedon, P.J. Piscaer, K. Batselier, C.S. Smith, M.H.G. Verhaegen
DOI related publication
https://doi.org/10.23919/EUSIPCO.2019.8902976
More Info
expand_more
Publication Year
2019
Language
English
Copyright
© 2019 Daniel Gedon, P.J. Piscaer, K. Batselier, C.S. Smith, M.H.G. Verhaegen
Research Group
Team Raf Van de Plas
ISBN (electronic)
978-9-0827-9703-9
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

An extension of the Tensor Network (TN) Kalman filter [2], [3] for large scale LTI systems is presented in this paper. The TN Kalman filter can handle exponentially large state vectors without constructing them explicitly. In order to have efficient algebraic operations, a low TN rank is required. We exploit the possibility to approximate the covariance matrix as a TN with a low TN rank. This reduces the computational complexity for general SISO and MIMO LTI systems with TN rank greater than one significantly while obtaining an accurate estimation. Improvements of this method in terms of computational complexity compared to the conventional Kalman filter are demonstrated in numerical simulations for large scale systems.

Files

08902976.pdf
(pdf | 0.559 Mb)
- Embargo expired in 18-05-2020
License info not available