Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes

The control of complex systems can be done decomposing the control task into a sequence of control modes, or modes for short. Each mode implements a parameterized feedback law until a termination condition is activated in response to the occurrence of an exogenous/endogenous event, which indicates t...

Full description

Bibliographic Details
Main Authors:	Mariano De Paula, Luis O. Ávila, Carlos sánchez Reinoso, Gerardo G. Acosta
Format:	Article
Language:	Spanish
Published:	Universitat Politecnica de Valencia 2015-10-01
Series:	Revista Iberoamericana de Automática e Informática Industrial RIAI
Subjects:	Control multimodal Programación dinámica Procesos Gaussianos Incertidumbre Política
Online Access:	https://polipapers.upv.es/index.php/RIAI/article/view/9340

id	doaj-2bbe58032194406cab0d01f3cf1cc7e0
record_format	Article
spelling	doaj-2bbe58032194406cab0d01f3cf1cc7e02021-04-02T14:03:31ZspaUniversitat Politecnica de ValenciaRevista Iberoamericana de Automática e Informática Industrial RIAI1697-79121697-79202015-10-0112438539610.1016/j.riai.2015.09.0046389Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian ProcessesMariano De Paula0Luis O. Ávila1Carlos sánchez Reinoso2Gerardo G. Acosta3Universidad Nacional del Centro de la Provincia de Buenos AiresCONICET-UTNUniversidad Nacional de Catamarca-CONICETUniversidad Nacional del Centro de la Provincia de Buenos AiresThe control of complex systems can be done decomposing the control task into a sequence of control modes, or modes for short. Each mode implements a parameterized feedback law until a termination condition is activated in response to the occurrence of an exogenous/endogenous event, which indicates that the execution mode must end. This paper presents a novel approach to find an optimal switching policy to solve a control problem by optimizing some measure of cost/benefit. An optimal policy implements an optimal multimodal control program, consisting in a sequence of control modes. The proposal includes the development of an algorithm based on the idea of dynamic programming integrating Gaussian processes and Bayesian active learning. In addition, an efficient use of the data to improve the exploration of the continuous state spaces solutions can be achieved through this approach. A representative case study is discussed and analyzed to demonstrate the performance of the proposed algorithm.https://polipapers.upv.es/index.php/RIAI/article/view/9340Control multimodalProgramación dinámicaProcesos GaussianosIncertidumbrePolítica
collection	DOAJ
language	Spanish
format	Article
sources	DOAJ
author	Mariano De Paula Luis O. Ávila Carlos sánchez Reinoso Gerardo G. Acosta
spellingShingle	Mariano De Paula Luis O. Ávila Carlos sánchez Reinoso Gerardo G. Acosta Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes Revista Iberoamericana de Automática e Informática Industrial RIAI Control multimodal Programación dinámica Procesos Gaussianos Incertidumbre Política
author_facet	Mariano De Paula Luis O. Ávila Carlos sánchez Reinoso Gerardo G. Acosta
author_sort	Mariano De Paula
title	Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes
title_short	Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes
title_full	Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes
title_fullStr	Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes
title_full_unstemmed	Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes
title_sort	multimodal control in uncertain environments using reinforcement learning and gaussian processes
publisher	Universitat Politecnica de Valencia
series	Revista Iberoamericana de Automática e Informática Industrial RIAI
issn	1697-7912 1697-7920
publishDate	2015-10-01
description	The control of complex systems can be done decomposing the control task into a sequence of control modes, or modes for short. Each mode implements a parameterized feedback law until a termination condition is activated in response to the occurrence of an exogenous/endogenous event, which indicates that the execution mode must end. This paper presents a novel approach to find an optimal switching policy to solve a control problem by optimizing some measure of cost/benefit. An optimal policy implements an optimal multimodal control program, consisting in a sequence of control modes. The proposal includes the development of an algorithm based on the idea of dynamic programming integrating Gaussian processes and Bayesian active learning. In addition, an efficient use of the data to improve the exploration of the continuous state spaces solutions can be achieved through this approach. A representative case study is discussed and analyzed to demonstrate the performance of the proposed algorithm.
topic	Control multimodal Programación dinámica Procesos Gaussianos Incertidumbre Política
url	https://polipapers.upv.es/index.php/RIAI/article/view/9340
work_keys_str_mv	AT marianodepaula multimodalcontrolinuncertainenvironmentsusingreinforcementlearningandgaussianprocesses AT luisoavila multimodalcontrolinuncertainenvironmentsusingreinforcementlearningandgaussianprocesses AT carlossanchezreinoso multimodalcontrolinuncertainenvironmentsusingreinforcementlearningandgaussianprocesses AT gerardogacosta multimodalcontrolinuncertainenvironmentsusingreinforcementlearningandgaussianprocesses
_version_	1721563140922015744

Multimodal Control in Uncertain Environments using Reinforcement Learning and Gaussian Processes

Similar Items