A4 Conference proceedings

Adaptive Square-Root Unscented Kalman Filter: Implementation of Exponential Forgetting Factor


Publication Details
Authors: Mohammadi Asl Reza, Shabbouei Hagh Yashar, Fekih Afef, Handroos Heikki
Publication year: 2020
Language: English
Title of parent publication: 2020 6th International Conference on Control, Automation and Robotics (ICCAR)
ISBN: 978-1-7281-6140-2
eISBN: 978-1-7281-6139-6
ISSN: 2251-2446
JUFO-Level of this publication: 1
Open Access: Not an Open Access publication

Abstract

This paper proposes a new form of adaptive square root unscented Kalman filter that implements an exponential forgetting factor to update the filter. It aims at estimating the states of nonlinear systems without a priori knowledge about the statistics of noises. The filter updates the estimation of covariances of noises with time, and the updated covariances are used to update the states of the system. The proposed approach is implemented to a servo-hydraulic system which states and measurements are affected by time varying noises with time-varying statistics. The obtained results along with the mean square errors of the estimation of states confirmed the performance and precision of the proposed filter.


Last updated on 2020-23-10 at 13:19