Control of a two-dimensional magnetic positioning system with deep reinforcement learning and feedback linearization

Eduardo Bejar, Antonio Moran

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

This paper presents a neuro-controller based on deep reinforcement learning to control the nonlinear dynamics of a two-dimensional magnetic positioning system. The feedback-linearized model of the magnetic positioning system is used to generate training data for the neuro-controller. The neuro-controller is trained using the Deep Deterministic Policy Gradient (DDPG) algorithm. The effectiveness of the proposed control strategy is verified with different desired setpoints and trajectories, and diverse working conditions.

Original languageEnglish
Title of host publication2018 IEEE 61st International Midwest Symposium on Circuits and Systems, MWSCAS 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages909-912
Number of pages4
ISBN (Electronic)9781538673928
DOIs
StatePublished - 2 Jul 2018
Event61st IEEE International Midwest Symposium on Circuits and Systems, MWSCAS 2018 - Windsor, Canada
Duration: 5 Aug 20188 Aug 2018

Publication series

NameMidwest Symposium on Circuits and Systems
Volume2018-August
ISSN (Print)1548-3746

Conference

Conference61st IEEE International Midwest Symposium on Circuits and Systems, MWSCAS 2018
Country/TerritoryCanada
CityWindsor
Period5/08/188/08/18

Keywords

  • Deep deterministic policy gradient algorithm (DDPG)
  • Magnetic positioning system
  • Q-learning
  • Reinforcement learning

Fingerprint

Dive into the research topics of 'Control of a two-dimensional magnetic positioning system with deep reinforcement learning and feedback linearization'. Together they form a unique fingerprint.

Cite this