Aprendizaje de movimientos en robot humanoide a partir de inferencia de objetivos

Suarez Huertas, Yeison Estiven

Aprendizaje de movimientos en robot humanoide a partir de inferencia de objetivos

dc.contributor.advisor	Camacho Poveda, Edgar Camilo	spa
dc.contributor.advisor	Higuera Arias, Carolina	spa
dc.contributor.author	Suarez Huertas, Yeison Estiven	spa
dc.contributor.cvlac	https://scienti.minciencias.gov.co/cvlac/visualizador/generarCurriculoCv.do?cod_rh=0001630084	spa
dc.contributor.cvlac	http://scienti.colciencias.gov.co:8081/cvlac/visualizador/generarCurriculoCv.do?cod_rh=	spa
dc.contributor.googlescholar	https://scholar.google.es/citations?user=tJG988kAAAAJ&hl=es	spa
dc.contributor.googlescholar	https://scholar.google.es/citations?user=ZaxycbsAAAAJ&hl=es	spa
dc.contributor.orcid	https://orcid.org/0000-0002-6084-2512	spa
dc.contributor.orcid	https://orcid.org/0000-0001-5141-0817	spa
dc.coverage.campus	CRAI-USTA Bogotá	spa
dc.date.accessioned	2020-09-26T00:10:16Z	spa
dc.date.available	2020-09-26T00:10:16Z	spa
dc.date.issued	2020-08-16	spa
dc.description	Este documento presenta la aplicación del aprendizaje por refuerzo inverso (IRL por sus siglas en inglés) en un robot humanoide conocido como Poppy Torso, con el fin de realizar movimientos de las extremidades superiores. El aprendizaje por refuerzo inverso se basa en el aprendizaje a partir de las demostraciones (trayectorias) de un experto. Con el fin de obtener una utilidad final lo más cercana a la utilidad obtenida por el experto en su recorrido, previamente se implementa un aprendizaje por refuerzo (RL por sus siglas en inglés) con una recompensa plenamente establecida dentro del entorno diseñado, el cual logró cumplir el objetivo que corresponde a generar movimientos desde un punto aleatorio hasta un punto establecido. El robot en simulación logra en la mayoría de los casos (con un porcentaje del 97.5% realizado sobre 1000 pruebas) llegar a su objetivo, tanto por aprendizaje por refuerzo como por refuerzo inverso.	spa
dc.description.abstract	This document presents the application of Inverse Reinforcement Learning (IRL) in a humanoid robot known as Poppy Torso, in order to perform upper extremity movements. Inverse Reinforcement Learning (IRL) is based on learning from the demonstrations (trajectories) of an expert. In order to obtain a final utility as close to the utility obtained by the expert in his task, reinforcement learning (RL) is previously implemented with a fully established reward within the designed environment, which achieved fulfill the objective that corresponds to generating movements from a random point to a set point. The robot in simulation achieves in most cases (with a percentage of 97.5% performed on 1000 tests) to reach its objective, both by reinforcement learning and by inverse reinforcement.	spa
dc.description.degreelevel	Pregrado	spa
dc.description.degreename	Ingeniero Electronico	spa
dc.description.domain	http://unidadinvestigacion.usta.edu.co	spa
dc.format.mimetype	application/pdf	spa
dc.identifier.citation	Suarez Huertas, Y. E. (2020). Aprendizaje de movimientos en robot humanoide a partir de inferencia de objetivos [tesis de pregrado, Universidad Santo Tomás] Repositorio instituconal - Universidad Santo Tomás	spa
dc.identifier.instname	instname:Universidad Santo Tomás	spa
dc.identifier.reponame	reponame:Repositorio Institucional Universidad Santo Tomás	spa
dc.identifier.repourl	repourl:https://repository.usta.edu.co	spa
dc.identifier.uri	http://hdl.handle.net/11634/30072
dc.language.iso	spa	spa
dc.publisher	Universidad Santo Tomás	spa
dc.publisher.faculty	Facultad de Ingeniería Electrónica	spa
dc.publisher.program	Pregrado Ingeniería Electrónica	spa
dc.relation.references	R. S. Sutton and A. G. Barto, Reinforcement Learning: An Introduction. The MIT Press, second ed., 1998.	spa
dc.relation.references	C. H. Yeison Suarez and E. C. Camacho, “Inverse reinforcement learning application for discrete and continuous environments,” in AETA 2019 - Recent Advances in Electrical Engineering and Related Sciences: Theory and Application, Springer International Publishing, 2020.	spa
dc.relation.references	J. G. Matthieu Lapeyre, Pierre Rouanet, “The poppy project.” https: //www.poppy-project.org/, 2012.	spa
dc.relation.references	P. Manceron, “Ikpy library.” https://github.com/Phylliade/ikpy, 2018.	spa
dc.relation.references	P. Abbeel and A. Y. Ng, ““apprenticeship learning via inverse reinforcement learning, ” in Proceedings of the Twenty-first International Conference on Machine Learning, ICML ’04, (New York, NY, USA),” 2004.	spa
dc.relation.references	H. van Hasselt, A. Guez, and D. Silver, “Deep reinforcement learning with double q-learning.” https://arxiv.org/pdf/1509.06461.pdf, 2015.	spa
dc.relation.references	D. S. A. G. I. A. Volodymyr Mnih, Koray Kavukcuoglu, “Playing atari with deep reinforcement learning.” https://arxiv.org/pdf/1312. 5602.pdf, 2013.	spa
dc.relation.references	N. Ratliff, J. A. Bagnell, and S. S. Srinivasa, “Imitation learning for locomotion and manipulation, ”in 2007 7th IEEE-RAS International Conference on Humanoid Robots. Nov. 2007.	spa
dc.relation.references	A. J. Ijspeert, J. Nakanishi, and S. Schaal, ““movement imitation with nonlinear dynamical systems in humanoid robots, ” in Proceedings 2002 IEEE International Conference on Robotics and Automation,” vol. 2, p. 1398–1403, May 2002.	spa
dc.relation.references	K. Mu¨lling, J. Kober, O. Kroemer, and J. Peters, “Learning to select and generalize striking movements in robot table tennis,” The International Journal of Robotics Research, vol. 32, no. 3, p. 263–279, 2013.	spa
dc.relation.references	B. D. Argall, S. Chernova, M. Veloso, and B. Browning, ““a survey of robot learning from demonstration, ” Robotics and Autonomous Systems,” vol. 57, no. 5, p. 469, 2009.	spa
dc.relation.references	S. Calinon, F. Guenter, and A. Billard, ““on learning, representing, and generalizing a task in a humanoid robot, ” IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics),” vol. 37, p. 286–298, Apr. 2007.	spa
dc.relation.references	P. Pastor, H. Hoffmann, T. Asfour, and S. Schaal, “Learning and generalization of motor skills by learning from demonstration, ” in 2009 IEEE International Conference on Robotics and Automation. May 2009.	spa
dc.relation.references	T. Zhang, Z. McCarthy, O. Jow, D. Lee, K. Goldberg, and P. Abbeel, ““deep imitation learning for complex manipulation tasks from virtual reality teleoperation, ” CoRR,” vol. abs/1710.04615, 2017.	spa
dc.relation.references	B. C. Stadie, P. Abbeel, and I. Sutskever, “Third-Person Imitation Learning, ” ArXiv e-prints. Mar. 2017.	spa
dc.relation.references	P. Sermanet, K. Xu, and S. Levine, ““unsupervised perceptual rewards for imitation learning, ” CoRR,” vol. abs/1612.06699, 2017.	spa
dc.relation.references	P. Sermanet, C. Lynch, Y. Chebotar, J. Hsu, E. Jang, S. Schaal, and S. Levine, “Time-contrastivenetworks: Self-supervised learning from video. 2018.	spa
dc.relation.references	B. Piot, M. Geist, and O. Pietquin, ““bridging the gap between imitation learning and inverse reinforcement learning,” vol. 28, p. 1814–1826, Aug. 2017.	spa
dc.relation.references	D. Silver, A. Huang, C. J. Maddison, A. Guez, L. Sifre, G. van den Driessche, J. Schrittwieser, I. Antonoglou, V. Panneershelvam, M. Lanctot, S. Dieleman, D. Grewe, J. Nham, N. Kalchbrenner, I. Sutskever, T. Lillicrap, M. Leach, K. Kavukcuoglu, T. Graepel, and D. Hassabis, “Mastering the game of go with deep neural networks and tree search,” Nature, vol. 529, pp. 484 EP –, Jan 2016. Article.	spa
dc.relation.references	T. Hester, M. Vecerik, O. Pietquin, M. Lanctot, T. Schaul, B. Piot, D. Horgan, J. Quan, A. Sendonaris, I. Osband, G. Dulac-Arnold, J. Agapiou, J. Z. Leibo, and A. Gruslys, Deep q-learning from demonstrations. feb 2018.	spa
dc.relation.references	G. Masuyama and K. Umeda, “Apprenticeship learning based on inconsistent demonstrations, ” in 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 2015.	spa
dc.relation.references	I. A. D. S. Tom Schaul, John Quan, “Prioritized experience replay.” https://arxiv.org/pdf/1511.05952.pdf, 2016.	spa
dc.rights	Atribución-NoComercial 2.5 Colombia	*
dc.rights	Atribución-NoComercial 2.5 Colombia	*
dc.rights.accessrights	info:eu-repo/semantics/openAccess
dc.rights.coar	http://purl.org/coar/access_right/c_abf2
dc.rights.local	Abierto (Texto Completo)	spa
dc.rights.uri	http://creativecommons.org/licenses/by-nc/2.5/co/	*
dc.subject.keyword	Reinforcement learning	spa
dc.subject.keyword	Inverse reinforcement learning	spa
dc.subject.keyword	Computational neural networks	spa
dc.subject.keyword	Machine Learning	spa
dc.subject.keyword	Python	spa
dc.subject.keyword	Poppy torso	spa
dc.subject.keyword	Inference objectives	spa
dc.subject.keyword	humanoid robot	spa
dc.subject.lemb	Redes neuronales computacionales	spa
dc.subject.lemb	Aprendizaje de máquina	spa
dc.subject.lemb	Inferencia de objetivos	spa
dc.subject.lemb	Humanoides	spa
dc.subject.proposal	Aprendizaje por refuerzo	spa
dc.subject.proposal	Aprendizaje por refuerzo inverso	spa
dc.subject.proposal	Python	spa
dc.subject.proposal	Poppy Torso	spa
dc.subject.proposal	Inferencia de objetivos	spa
dc.subject.proposal	Robot Humanoide	spa
dc.title	Aprendizaje de movimientos en robot humanoide a partir de inferencia de objetivos	spa
dc.type	bachelor thesis
dc.type.category	Formación de Recurso Humano para la Ctel: Trabajo de grado de Pregrado	spa
dc.type.coar	http://purl.org/coar/resource_type/c_7a1f
dc.type.coarversion	http://purl.org/coar/version/c_ab4af688f83e57aa
dc.type.drive	info:eu-repo/semantics/bachelorThesis
dc.type.local	Tesis de pregrado	spa
dc.type.version	info:eu-repo/semantics/acceptedVersion