Print Email Facebook Twitter Reinforcement Learning for Orientation Estimation Using Inertial Sensors with Performance Guarantee Title Reinforcement Learning for Orientation Estimation Using Inertial Sensors with Performance Guarantee Author Hu, Liang (University of Essex) Tang, Y. (TU Delft Robot Dynamics) Zhou, Z. (TU Delft Robot Dynamics) Pan, W. (TU Delft Robot Dynamics) Date 2021 Abstract This paper presents a deep reinforcement learning (DRL) algorithm for orientation estimation using inertial sensors combined with a magnetometer. Lyapunov’s method in control theory is employed to prove the convergence of orientation estimation errors. The estimator gains and a Lyapunov function are parametrised by deep neural networks and learned from samples based on the theoretical results. The DRL estimator is compared with three well-known orientation estimation methods on both numerical simulations and real dataset collected from commercially available sensors. The results show that the proposed algorithm is superior for arbitrary estimation initialisation and can adapt to a drastic angular velocity profile for which other algorithms can be hardly applicable. To the best of our knowledge, this is the first DRL-based orientation estimation method with an estimation error boundedness guarantee. To reference this document use: http://resolver.tudelft.nl/uuid:358694af-f535-48be-b7d6-ba4fba1c2fcc DOI https://doi.org/10.1109/ICRA48506.2021.9561440 Publisher IEEE ISBN 978-1-7281-9078-5 Source Proceedings of the 2021 IEEE International Conference on Robotics and Automation (ICRA) Event ICRA 2021, 2021-05-30 → 2021-06-05, Hybrid at Xi'an, China Bibliographical note Accepted Author Manuscript Part of collection Institutional Repository Document type conference paper Rights © 2021 Liang Hu, Y. Tang, Z. Zhou, W. Pan Files PDF Reinforcement_Learning_fo ... rantee.pdf 6.14 MB Close viewer /islandora/object/uuid:358694af-f535-48be-b7d6-ba4fba1c2fcc/datastream/OBJ/view