Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation

Yujie Tang; Liang  Hu; Qingrui Zhang; Wei Pan

doi:10.1109/IROS51168.2021.9635963

Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation

Yujie Tang, Liang Hu, Qingrui Zhang, Wei Pan^*

^*Corresponding author for this work

Robot Dynamics

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

7 Citations (Scopus)

Abstract

Inertial measurement units are widely used in different fields to estimate the attitude. Many algorithms have been proposed to improve estimation performance. However, most of them still suffer from 1) inaccurate initial estimation, 2) inaccurate initial filter gain, and 3) non-Gaussian process and/or measurement noise. This paper will leverage reinforcement learning to compensate for the classical extended Kalman filter estimation, i.e., to learn the filter gain from the sensor measurements. We also analyse the convergence of the estimate error. The effectiveness of the proposed algorithm is validated on both simulated data and real data.

Original language	English
Title of host publication	Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021)
Publisher	IEEE
Pages	6854-6859
ISBN (Electronic)	978-1-6654-1714-3
ISBN (Print)	978-1-6654-1715-0
DOIs	https://doi.org/10.1109/IROS51168.2021.9635963
Publication status	Published - 2021
Event	2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) - Online at Prague, Czech Republic Duration: 27 Sept 2021 → 1 Oct 2021

Conference

Conference	2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Country/Territory	Czech Republic
City	Online at Prague
Period	27/09/21 → 1/10/21

Access to Document

10.1109/IROS51168.2021.9635963

Cite this

@inproceedings{affec45626a54f939bd35e4cdb1be768,

title = "Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation",

abstract = "Inertial measurement units are widely used in different fields to estimate the attitude. Many algorithms have been proposed to improve estimation performance. However, most of them still suffer from 1) inaccurate initial estimation, 2) inaccurate initial filter gain, and 3) non-Gaussian process and/or measurement noise. This paper will leverage reinforcement learning to compensate for the classical extended Kalman filter estimation, i.e., to learn the filter gain from the sensor measurements. We also analyse the convergence of the estimate error. The effectiveness of the proposed algorithm is validated on both simulated data and real data.",

author = "Yujie Tang and Liang Hu and Qingrui Zhang and Wei Pan",

year = "2021",

doi = "10.1109/IROS51168.2021.9635963",

language = "English",

isbn = "978-1-6654-1715-0",

pages = "6854--6859",

booktitle = "Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021)",

publisher = "IEEE",

address = "United States",

note = "2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) ; Conference date: 27-09-2021 Through 01-10-2021",

}

Tang, Y, Hu, L, Zhang, Q & Pan, W 2021, Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation. in Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021). IEEE, pp. 6854-6859, 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Online at Prague, Czech Republic, 27/09/21. https://doi.org/10.1109/IROS51168.2021.9635963

Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation. / Tang, Yujie; Hu, Liang ; Zhang, Qingrui et al.
Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021). IEEE, 2021. p. 6854-6859.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation

AU - Tang, Yujie

AU - Hu, Liang

AU - Zhang, Qingrui

AU - Pan, Wei

PY - 2021

Y1 - 2021

N2 - Inertial measurement units are widely used in different fields to estimate the attitude. Many algorithms have been proposed to improve estimation performance. However, most of them still suffer from 1) inaccurate initial estimation, 2) inaccurate initial filter gain, and 3) non-Gaussian process and/or measurement noise. This paper will leverage reinforcement learning to compensate for the classical extended Kalman filter estimation, i.e., to learn the filter gain from the sensor measurements. We also analyse the convergence of the estimate error. The effectiveness of the proposed algorithm is validated on both simulated data and real data.

AB - Inertial measurement units are widely used in different fields to estimate the attitude. Many algorithms have been proposed to improve estimation performance. However, most of them still suffer from 1) inaccurate initial estimation, 2) inaccurate initial filter gain, and 3) non-Gaussian process and/or measurement noise. This paper will leverage reinforcement learning to compensate for the classical extended Kalman filter estimation, i.e., to learn the filter gain from the sensor measurements. We also analyse the convergence of the estimate error. The effectiveness of the proposed algorithm is validated on both simulated data and real data.

UR - http://www.scopus.com/inward/record.url?scp=85124368470&partnerID=8YFLogxK

U2 - 10.1109/IROS51168.2021.9635963

DO - 10.1109/IROS51168.2021.9635963

M3 - Conference contribution

SN - 978-1-6654-1715-0

SP - 6854

EP - 6859

BT - Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021)

PB - IEEE

T2 - 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Y2 - 27 September 2021 through 1 October 2021

ER -

Reinforcement Learning Compensated Extended Kalman Filter for Attitude Estimation

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this