Automated design of complex dynamic systems.

Several fields of study are concerned with uniting the concept of computation with that of the design of physical systems. For example, a recent trend in robotics is to design robots in such a way that they require a minimal control effort. Another example is found in the domain of photonics, where...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Michiel Hermans, Benjamin Schrauwen, Peter Bienstman, Joni Dambre
Formato:	article
Lenguaje:	EN
Publicado:	Public Library of Science (PLoS) 2014
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/32eda949ce234cb8b9ddb05e20a32399
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:32eda949ce234cb8b9ddb05e20a32399
record_format	dspace
spelling	oai:doaj.org-article:32eda949ce234cb8b9ddb05e20a323992021-11-18T08:34:40ZAutomated design of complex dynamic systems.1932-620310.1371/journal.pone.0086696https://doaj.org/article/32eda949ce234cb8b9ddb05e20a323992014-01-01T00:00:00Zhttps://www.ncbi.nlm.nih.gov/pmc/articles/pmid/24497969/pdf/?tool=EBIhttps://doaj.org/toc/1932-6203Several fields of study are concerned with uniting the concept of computation with that of the design of physical systems. For example, a recent trend in robotics is to design robots in such a way that they require a minimal control effort. Another example is found in the domain of photonics, where recent efforts try to benefit directly from the complex nonlinear dynamics to achieve more efficient signal processing. The underlying goal of these and similar research efforts is to internalize a large part of the necessary computations within the physical system itself by exploiting its inherent non-linear dynamics. This, however, often requires the optimization of large numbers of system parameters, related to both the system's structure as well as its material properties. In addition, many of these parameters are subject to fabrication variability or to variations through time. In this paper we apply a machine learning algorithm to optimize physical dynamic systems. We show that such algorithms, which are normally applied on abstract computational entities, can be extended to the field of differential equations and used to optimize an associated set of parameters which determine their behavior. We show that machine learning training methodologies are highly useful in designing robust systems, and we provide a set of both simple and complex examples using models of physical dynamical systems. Interestingly, the derived optimization method is intimately related to direct collocation a method known in the field of optimal control. Our work suggests that the application domains of both machine learning and optimal control have a largely unexplored overlapping area which envelopes a novel design methodology of smart and highly complex physical systems.Michiel HermansBenjamin SchrauwenPeter BienstmanJoni DambrePublic Library of Science (PLoS)articleMedicineRScienceQENPLoS ONE, Vol 9, Iss 1, p e86696 (2014)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Michiel Hermans Benjamin Schrauwen Peter Bienstman Joni Dambre Automated design of complex dynamic systems.
description	Several fields of study are concerned with uniting the concept of computation with that of the design of physical systems. For example, a recent trend in robotics is to design robots in such a way that they require a minimal control effort. Another example is found in the domain of photonics, where recent efforts try to benefit directly from the complex nonlinear dynamics to achieve more efficient signal processing. The underlying goal of these and similar research efforts is to internalize a large part of the necessary computations within the physical system itself by exploiting its inherent non-linear dynamics. This, however, often requires the optimization of large numbers of system parameters, related to both the system's structure as well as its material properties. In addition, many of these parameters are subject to fabrication variability or to variations through time. In this paper we apply a machine learning algorithm to optimize physical dynamic systems. We show that such algorithms, which are normally applied on abstract computational entities, can be extended to the field of differential equations and used to optimize an associated set of parameters which determine their behavior. We show that machine learning training methodologies are highly useful in designing robust systems, and we provide a set of both simple and complex examples using models of physical dynamical systems. Interestingly, the derived optimization method is intimately related to direct collocation a method known in the field of optimal control. Our work suggests that the application domains of both machine learning and optimal control have a largely unexplored overlapping area which envelopes a novel design methodology of smart and highly complex physical systems.
format	article
author	Michiel Hermans Benjamin Schrauwen Peter Bienstman Joni Dambre
author_facet	Michiel Hermans Benjamin Schrauwen Peter Bienstman Joni Dambre
author_sort	Michiel Hermans
title	Automated design of complex dynamic systems.
title_short	Automated design of complex dynamic systems.
title_full	Automated design of complex dynamic systems.
title_fullStr	Automated design of complex dynamic systems.
title_full_unstemmed	Automated design of complex dynamic systems.
title_sort	automated design of complex dynamic systems.
publisher	Public Library of Science (PLoS)
publishDate	2014
url	https://doaj.org/article/32eda949ce234cb8b9ddb05e20a32399
work_keys_str_mv	AT michielhermans automateddesignofcomplexdynamicsystems AT benjaminschrauwen automateddesignofcomplexdynamicsystems AT peterbienstman automateddesignofcomplexdynamicsystems AT jonidambre automateddesignofcomplexdynamicsystems
_version_	1718421612475187200

Automated design of complex dynamic systems.

Ejemplares similares