Actor-critic reinforcement learning for tracking control in robotics

Yudha P. Pane; Subramanya P. Nageshrao; Robert Babuska

doi:10.1109/CDC.2016.7799164

Actor-critic reinforcement learning for tracking control in robotics

Yudha P. Pane, Subramanya P. Nageshrao, Robert Babuska

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

33 Citations (SciVal)

Abstract

In this article we provide experimental results and evaluation of a compensation method which improves the tracking performance of a nominal feedback controller by means of reinforcement learning (RL). The compensator is based on the actor-critic scheme and it adds a correction signal to the nominal control input with the goal to improve the tracking performance using on-line learning. The algorithm has been evaluated on a 6 DOF industrial robot manipulator with the objective to accurately track different types of reference trajectories. An extensive experimental study has shown that the proposed RL-based compensation method significantly improves the performance of the nominal feedback controller.

Original language	English
Title of host publication	Proceedings 2016 IEEE 55th Conference on Decision and Control (CDC)
Editors	Francesco Bullo, Christophe Prieur, Alessandro Giua
Place of Publication	Piscataway, NJ, USA
Publisher	IEEE
Pages	5819-5826
ISBN (Electronic)	978-1-5090-1837-6
DOIs	https://doi.org/10.1109/CDC.2016.7799164
Publication status	Published - 2016
Event	55th IEEE Conference on Decision and Control, CDC 2016 - Las Vegas, United States Duration: 12 Dec 2016 → 14 Dec 2016

Conference

Conference	55th IEEE Conference on Decision and Control, CDC 2016
Abbreviated title	CDC 2016
Country/Territory	United States
City	Las Vegas
Period	12/12/16 → 14/12/16

Access to Document

10.1109/CDC.2016.7799164

Cite this

@inproceedings{54c3bfa863734e46a7e144cb7deaeed6,

title = "Actor-critic reinforcement learning for tracking control in robotics",

abstract = "In this article we provide experimental results and evaluation of a compensation method which improves the tracking performance of a nominal feedback controller by means of reinforcement learning (RL). The compensator is based on the actor-critic scheme and it adds a correction signal to the nominal control input with the goal to improve the tracking performance using on-line learning. The algorithm has been evaluated on a 6 DOF industrial robot manipulator with the objective to accurately track different types of reference trajectories. An extensive experimental study has shown that the proposed RL-based compensation method significantly improves the performance of the nominal feedback controller.",

author = "Pane, \{Yudha P.\} and Nageshrao, \{Subramanya P.\} and Robert Babuska",

year = "2016",

doi = "10.1109/CDC.2016.7799164",

language = "English",

pages = "5819--5826",

editor = "Francesco Bullo and Christophe Prieur and Alessandro Giua",

booktitle = "Proceedings 2016 IEEE 55th Conference on Decision and Control (CDC)",

publisher = "IEEE",

address = "United States",

note = "55th IEEE Conference on Decision and Control, CDC 2016, CDC 2016 ; Conference date: 12-12-2016 Through 14-12-2016",

}

Pane, YP, Nageshrao, SP & Babuska, R 2016, Actor-critic reinforcement learning for tracking control in robotics. in F Bullo, C Prieur & A Giua (eds), Proceedings 2016 IEEE 55th Conference on Decision and Control (CDC) . IEEE, Piscataway, NJ, USA, pp. 5819-5826, 55th IEEE Conference on Decision and Control, CDC 2016, Las Vegas, United States, 12/12/16. https://doi.org/10.1109/CDC.2016.7799164

Actor-critic reinforcement learning for tracking control in robotics. / Pane, Yudha P.; Nageshrao, Subramanya P.; Babuska, Robert.
Proceedings 2016 IEEE 55th Conference on Decision and Control (CDC) . ed. / Francesco Bullo; Christophe Prieur; Alessandro Giua. Piscataway, NJ, USA: IEEE, 2016. p. 5819-5826.

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

TY - GEN

T1 - Actor-critic reinforcement learning for tracking control in robotics

AU - Pane, Yudha P.

AU - Nageshrao, Subramanya P.

AU - Babuska, Robert

PY - 2016

Y1 - 2016

N2 - In this article we provide experimental results and evaluation of a compensation method which improves the tracking performance of a nominal feedback controller by means of reinforcement learning (RL). The compensator is based on the actor-critic scheme and it adds a correction signal to the nominal control input with the goal to improve the tracking performance using on-line learning. The algorithm has been evaluated on a 6 DOF industrial robot manipulator with the objective to accurately track different types of reference trajectories. An extensive experimental study has shown that the proposed RL-based compensation method significantly improves the performance of the nominal feedback controller.

AB - In this article we provide experimental results and evaluation of a compensation method which improves the tracking performance of a nominal feedback controller by means of reinforcement learning (RL). The compensator is based on the actor-critic scheme and it adds a correction signal to the nominal control input with the goal to improve the tracking performance using on-line learning. The algorithm has been evaluated on a 6 DOF industrial robot manipulator with the objective to accurately track different types of reference trajectories. An extensive experimental study has shown that the proposed RL-based compensation method significantly improves the performance of the nominal feedback controller.

UR - https://www.scopus.com/pages/publications/85010734930

U2 - 10.1109/CDC.2016.7799164

DO - 10.1109/CDC.2016.7799164

M3 - Conference contribution

AN - SCOPUS:85010734930

SP - 5819

EP - 5826

BT - Proceedings 2016 IEEE 55th Conference on Decision and Control (CDC)

A2 - Bullo, Francesco

A2 - Prieur, Christophe

A2 - Giua, Alessandro

PB - IEEE

CY - Piscataway, NJ, USA

T2 - 55th IEEE Conference on Decision and Control, CDC 2016

Y2 - 12 December 2016 through 14 December 2016

ER -

Actor-critic reinforcement learning for tracking control in robotics

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this