Moving in a Simulated Environment Through Deep Reinforcement Learning

Reinforcement learning is a field of artificial intelligence that is continuously evolving and has a wide variety of applications. In recent years major progress has been made in the application of deep reinforcement learning to highdimensional problems with continuous state and action spaces. T...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Esarte, Javier, Folino, Pablo Daniel, Gómez, Juan Carlos
Formato: Documento de conferencia publisherVersion
Lenguaje:Inglés
Inglés
Publicado: IEEE 2024
Materias:
deep reinforcement learning
soft actor-critic
tetrapod robot
virtual environment
predictive control
machine learning
robotics
artificial neural networks
Acceso en línea:http://hdl.handle.net/20.500.12272/11143
Aporte de:
RIA - Repositorio Institucional Abierto (UTN) de Universidad Tecnológica Nacional
id I68-R174-20.500.12272-11143
record_format dspace
spelling I68-R174-20.500.12272-111432024-07-23T14:52:29Z Moving in a Simulated Environment Through Deep Reinforcement Learning Esarte, Javier Folino, Pablo Daniel Gómez, Juan Carlos deep reinforcement learning soft actor-critic tetrapod robot virtual environment predictive control machine learning robotics artificial neural networks Reinforcement learning is a field of artificial intelligence that is continuously evolving and has a wide variety of applications. In recent years major progress has been made in the application of deep reinforcement learning to highdimensional problems with continuous state and action spaces. This paper presents a complete analysis of the application of the soft actor-critic algorithm to teach a four legged robot with three joints on each leg how to move towards the center of a virtually simulated environment. The general formulation of the reinforcement learning problem is first presented, followed by the description of the environment under analysis and the applied algorithm. Afterwards, the obtained results are compared against those of a manually programmed policy, closing with a discussion of some key design choices and common challenges. UTN FRBA Convocatoria Viajes y Eventos FRBA año 2022 Esarte, Javier. Universidad Tecnológica Nacional. Facultad Regional de Buenos Aires. Grupo de Inteligencia Artificial y Robótica; Argentina. Folino, Pablo Daniel. Universidad Tecnológica Nacional. Facultad Regional de Buenos Aires. Grupo de Inteligencia Artificial y Robótica; Argentina. Gómez, Juan Carlos. Instituto Nacional de Tecnología Industrial. Grupo de Inteligencia Artificial; Argentina. 2024-07-23T14:52:28Z 2024-07-23T14:52:28Z 2022-09-08 info:eu-repo/semantics/conferenceObject publisherVersion J. Esarte, P. D. Folino and J. C. Gómez, "Moving in a Simulated Environment Through Deep Reinforcement Learning," 2022 IEEE Biennial Congress of Argentina (ARGENCON), San Juan, Argentina, 2022, pp. 1-6, doi: 10.1109/ARGENCON55245.2022.9939868. 978-1-6654-8014-7 978-1-6654-8015-4 http://hdl.handle.net/20.500.12272/11143 10.1109/ARGENCON55245.2022.9939868 eng eng openAccess Attribution-NonCommercial-NoDerivatives 4.0 Internacional http://creativecommons.org/licenses/by-nc-nd/4.0/ Attribution-NonCommercial-NoDerivatives 4.0 Internacional Atribución (“Creative Commons Attribution”): En cualquier explotación de la obra autorizada por la licencia será necesario reconocer la autoría (obligatoria en todos los casos). No comercial (“Creative Commons Non Commercial”): La explotación de la obra queda limitada a usos no comerciales. Sin obras derivadas (“Creative Commons No Derivate Works”): La autorización para explotar la obra no incluye la posibilidad de crear una obra derivada (traducciones, adaptaciones, etc.). pdf IEEE
institution Universidad Tecnológica Nacional
institution_str I-68
repository_str R-174
collection RIA - Repositorio Institucional Abierto (UTN)
language Inglés
Inglés
topic deep reinforcement learning
soft actor-critic
tetrapod robot
virtual environment
predictive control
machine learning
robotics
artificial neural networks
spellingShingle deep reinforcement learning
soft actor-critic
tetrapod robot
virtual environment
predictive control
machine learning
robotics
artificial neural networks
Esarte, Javier
Folino, Pablo Daniel
Gómez, Juan Carlos
Moving in a Simulated Environment Through Deep Reinforcement Learning
topic_facet deep reinforcement learning
soft actor-critic
tetrapod robot
virtual environment
predictive control
machine learning
robotics
artificial neural networks
description Reinforcement learning is a field of artificial intelligence that is continuously evolving and has a wide variety of applications. In recent years major progress has been made in the application of deep reinforcement learning to highdimensional problems with continuous state and action spaces. This paper presents a complete analysis of the application of the soft actor-critic algorithm to teach a four legged robot with three joints on each leg how to move towards the center of a virtually simulated environment. The general formulation of the reinforcement learning problem is first presented, followed by the description of the environment under analysis and the applied algorithm. Afterwards, the obtained results are compared against those of a manually programmed policy, closing with a discussion of some key design choices and common challenges.
format Documento de conferencia
publisherVersion
author Esarte, Javier
Folino, Pablo Daniel
Gómez, Juan Carlos
author_facet Esarte, Javier
Folino, Pablo Daniel
Gómez, Juan Carlos
author_sort Esarte, Javier
title Moving in a Simulated Environment Through Deep Reinforcement Learning
title_short Moving in a Simulated Environment Through Deep Reinforcement Learning
title_full Moving in a Simulated Environment Through Deep Reinforcement Learning
title_fullStr Moving in a Simulated Environment Through Deep Reinforcement Learning
title_full_unstemmed Moving in a Simulated Environment Through Deep Reinforcement Learning
title_sort moving in a simulated environment through deep reinforcement learning
publisher IEEE
publishDate 2024
url http://hdl.handle.net/20.500.12272/11143
work_keys_str_mv AT esartejavier movinginasimulatedenvironmentthroughdeepreinforcementlearning
AT folinopablodaniel movinginasimulatedenvironmentthroughdeepreinforcementlearning
AT gomezjuancarlos movinginasimulatedenvironmentthroughdeepreinforcementlearning
_version_ 1809230364598599680

Ejemplares similares