Levenberg-Marquardt and Line-Search Extended Kalman Smoothers
Paper in proceeding, 2020

The aim of this article is to present Levenberg–Marquardt and line-search extensions of the classical iterated extended Kalman smoother (IEKS) which has previously been shown to be equivalent to the Gauss–Newton method. The algo- rithms are derived by rewriting the algorithm’s steps in forms that can be efficiently implemented using modified EKS iter- ations. The resulting algorithms are experimentally shown to have superior convergence properties over the classical IEKS.

Levenberg–Marquardt algorithm

Extended Kalman smoother

nonlinear estimation

line search

Author

Simo Särkkä

Aalto University

Lennart Svensson

Chalmers, Electrical Engineering, Signal Processing and Biomedical Engineering

Published in

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

15206149 (ISSN)

Vol. 2020-May p. 5875-5879 art. no 9054686
9781509066315 (ISBN)

Conference

IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Barcelona / Online, Spain, 2020-05-03 - 2020-05-07

Research Project(s)

Deep multi-object tracking for ground truth trajectory estimation

VINNOVA (2017-05521), 2018-07-01 -- 2022-06-30.

Categorizing

Areas of Advance

Information and Communication Technology

Transport

Subject Categories (SSIF 2011)

Telecommunications

Probability Theory and Statistics

Signal Processing

Identifiers

DOI

10.1109/ICASSP40776.2020.9054686

More information

Latest update

4/5/2022 6