Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator

Brian R. Cairl, Farshad Khorrami

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The objective of this work is to achieve disturbance rejection and constant orientation of the trunk of a multi-legged robot. This is significant when payloads (such as cameras, optical systems, armaments) are carried by the robot. In particular, this paper presents an application of an on-line learning method to actively correct the open-loop gait generated by a central pattern generator (CPG) or a limit-cycle method. The learning method employed is based on a Nonlinear Autoregressive Neural Network with Exogenous inputs (NARX-NN)- a recurrent neural network architecture typically utilized for modeling nonlinear difference systems. A supervised learning approach is used to train the NARX-NN. The input to the neural network includes states of the robot legs, trunk attitude and attitude rates, and foot contact forces. The neural network is used to estimate the total torque imparted on the robot. The learned effects of the internal forces and disturbances are then applied in an inverse dynamics/computed torque controller, which is utilized to achieve a stable trunk (i.e., a constant orientation of the trunk). The efficacy of the proposed approach is shown in detailed simulation studies of a quadruped robot.

Original languageEnglish (US)
Title of host publicationIROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages6298-6303
Number of pages6
Volume2015-December
ISBN (Print)9781479999941
DOIs
StatePublished - Dec 11 2015
EventIEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2015 - Hamburg, Germany
Duration: Sep 28 2015Oct 2 2015

Other

OtherIEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2015
CountryGermany
CityHamburg
Period9/28/1510/2/15

Fingerprint

Stabilization
Robots
Neural networks
Torque
Disturbance rejection
Recurrent neural networks
Supervised learning
Network architecture
Optical systems
Cameras
Controllers

Keywords

  • Dynamics
  • Foot
  • Legged locomotion
  • Neural networks
  • Torque
  • Training

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Cairl, B. R., & Khorrami, F. (2015). Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator. In IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems (Vol. 2015-December, pp. 6298-6303). [7354276] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IROS.2015.7354276

Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator. / Cairl, Brian R.; Khorrami, Farshad.

IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2015-December Institute of Electrical and Electronics Engineers Inc., 2015. p. 6298-6303 7354276.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Cairl, BR & Khorrami, F 2015, Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator. in IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems. vol. 2015-December, 7354276, Institute of Electrical and Electronics Engineers Inc., pp. 6298-6303, IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2015, Hamburg, Germany, 9/28/15. https://doi.org/10.1109/IROS.2015.7354276
Cairl BR, Khorrami F. Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator. In IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2015-December. Institute of Electrical and Electronics Engineers Inc. 2015. p. 6298-6303. 7354276 https://doi.org/10.1109/IROS.2015.7354276
Cairl, Brian R. ; Khorrami, Farshad. / Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator. IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2015-December Institute of Electrical and Electronics Engineers Inc., 2015. pp. 6298-6303
@inproceedings{0bf1107ed11743c5ae09594772f6d427,
title = "Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator",
abstract = "The objective of this work is to achieve disturbance rejection and constant orientation of the trunk of a multi-legged robot. This is significant when payloads (such as cameras, optical systems, armaments) are carried by the robot. In particular, this paper presents an application of an on-line learning method to actively correct the open-loop gait generated by a central pattern generator (CPG) or a limit-cycle method. The learning method employed is based on a Nonlinear Autoregressive Neural Network with Exogenous inputs (NARX-NN)- a recurrent neural network architecture typically utilized for modeling nonlinear difference systems. A supervised learning approach is used to train the NARX-NN. The input to the neural network includes states of the robot legs, trunk attitude and attitude rates, and foot contact forces. The neural network is used to estimate the total torque imparted on the robot. The learned effects of the internal forces and disturbances are then applied in an inverse dynamics/computed torque controller, which is utilized to achieve a stable trunk (i.e., a constant orientation of the trunk). The efficacy of the proposed approach is shown in detailed simulation studies of a quadruped robot.",
keywords = "Dynamics, Foot, Legged locomotion, Neural networks, Torque, Training",
author = "Cairl, {Brian R.} and Farshad Khorrami",
year = "2015",
month = "12",
day = "11",
doi = "10.1109/IROS.2015.7354276",
language = "English (US)",
isbn = "9781479999941",
volume = "2015-December",
pages = "6298--6303",
booktitle = "IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Trunk stabilization of multi-legged robots using on-line learning via a NARX neural network compensator

AU - Cairl, Brian R.

AU - Khorrami, Farshad

PY - 2015/12/11

Y1 - 2015/12/11

N2 - The objective of this work is to achieve disturbance rejection and constant orientation of the trunk of a multi-legged robot. This is significant when payloads (such as cameras, optical systems, armaments) are carried by the robot. In particular, this paper presents an application of an on-line learning method to actively correct the open-loop gait generated by a central pattern generator (CPG) or a limit-cycle method. The learning method employed is based on a Nonlinear Autoregressive Neural Network with Exogenous inputs (NARX-NN)- a recurrent neural network architecture typically utilized for modeling nonlinear difference systems. A supervised learning approach is used to train the NARX-NN. The input to the neural network includes states of the robot legs, trunk attitude and attitude rates, and foot contact forces. The neural network is used to estimate the total torque imparted on the robot. The learned effects of the internal forces and disturbances are then applied in an inverse dynamics/computed torque controller, which is utilized to achieve a stable trunk (i.e., a constant orientation of the trunk). The efficacy of the proposed approach is shown in detailed simulation studies of a quadruped robot.

AB - The objective of this work is to achieve disturbance rejection and constant orientation of the trunk of a multi-legged robot. This is significant when payloads (such as cameras, optical systems, armaments) are carried by the robot. In particular, this paper presents an application of an on-line learning method to actively correct the open-loop gait generated by a central pattern generator (CPG) or a limit-cycle method. The learning method employed is based on a Nonlinear Autoregressive Neural Network with Exogenous inputs (NARX-NN)- a recurrent neural network architecture typically utilized for modeling nonlinear difference systems. A supervised learning approach is used to train the NARX-NN. The input to the neural network includes states of the robot legs, trunk attitude and attitude rates, and foot contact forces. The neural network is used to estimate the total torque imparted on the robot. The learned effects of the internal forces and disturbances are then applied in an inverse dynamics/computed torque controller, which is utilized to achieve a stable trunk (i.e., a constant orientation of the trunk). The efficacy of the proposed approach is shown in detailed simulation studies of a quadruped robot.

KW - Dynamics

KW - Foot

KW - Legged locomotion

KW - Neural networks

KW - Torque

KW - Training

UR - http://www.scopus.com/inward/record.url?scp=84958174831&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84958174831&partnerID=8YFLogxK

U2 - 10.1109/IROS.2015.7354276

DO - 10.1109/IROS.2015.7354276

M3 - Conference contribution

SN - 9781479999941

VL - 2015-December

SP - 6298

EP - 6303

BT - IROS Hamburg 2015 - Conference Digest: IEEE/RSJ International Conference on Intelligent Robots and Systems

PB - Institute of Electrical and Electronics Engineers Inc.

ER -