Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.13087/2675
Title: Improved particle filter-based estimation of a quadrotor subjected to uncertainties
Authors: Kaba, Aziz
Ermeydan, Ahmet
Keywords: UAV
Uncertainty
Estimation
Quadrotor
Particle filter
Resampling
Localization
Issue Date: 2022
Publisher: Emerald Group Publishing Ltd
Abstract: Purpose The purpose of this paper is to present an improved particle filter-based attitude estimator for a quadrotor unmanned aerial vehicle (UAV) that addresses the degeneracy issues. Design/methodology/approach Control of a quadrotor is not sufficient enough without an estimator to eliminate the noise from low-cost sensors. In this work, particle filter-based attitude estimator is proposed and used for nonlinear quadrotor dynamics. But, since recursive Bayesian estimation steps may rise degeneracy issues, the proposed scheme is improved with four different and widely used resampling algorithms. Findings Robustness of the proposed schemes is tested under various scenarios that include different levels of uncertainty and different particle sizes. Statistical analyses are conducted to assess the error performance of the schemes. According to the statistical analysis, the proposed estimators are capable of reducing sensor noise up to 5x, increasing signal to noise ratio up to 2.5x and reducing the uncertainty bounds up to 36x with root mean square value of as low as 0.0024, mean absolute error value of 0.036, respectively. Originality/value To the best of the authors' knowledge, the originality of this paper is to propose a robust particle filter-based attitude estimator to eliminate the low-cost sensor errors of quadrotor UAVs.
URI: https://doi.org/10.1108/AEAT-08-2021-0234
https://hdl.handle.net/20.500.13087/2675
ISSN: 1748-8842
1758-4213
Appears in Collections:Scopus İndeksli Yayınlar Koleksiyonu
WoS İndeksli Yayınlar Koleksiyonu

Show full item record

CORE Recommender

Page view(s)

6
checked on Oct 3, 2022

Google ScholarTM

Check

Altmetric


Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.