Aplicación de técnicas de Deep Learning en modelamiento y observación de la fotósfera solar

dc.contributor.advisorVargas Domínguez, Santiagospa
dc.contributor.advisorShelyag, Sergiyspa
dc.contributor.authorMorales Suarez, Germain Nicolasspa
dc.date.accessioned2024-04-29T19:54:42Z
dc.date.available2024-04-29T19:54:42Z
dc.date.issued2023
dc.descriptionilustraciones, diagramasspa
dc.description.abstractEl presente trabajo se enmarca en las aplicaciones de las redes neuronales profundas para el modelamiento de los fenómenos presentes en la fotósfera solar. La investigación propuesta se basa en la construcción de red neuronal convolucional 3D profunda de tipo generativa, DCGAN por sus siglas en ingles, haciendo uso de las módulos de inteligencia artificial de Python como Pytorch para arquitectura de la de red neuronal. Se pretende entrenar una red neuronal capaz de generar grupos de cubos de una alta similitud con cubos de entrenamiento, dichos cubos corresponden a magnitudes físicas de la fotósfera solar tales como densidad, campo magnético, velocidad del plasma, temperatura, entre otras, obtenidas del código de simulación MURaM. Codigo de simulación desarrollado por el grupo Solar-MHD de instituto Max Planck desarrollado entre el 2001-2005 con la finalidad de generar simulaciones realistas de procesos de magneto-convección y actividades magneticas, que tienen caso sobre la zona convectiva del sol, el presente trabajo busca tomar sus resultado y tomarlos como datos de entrenamiento para la red neuronal generando datos nuevos con una similitud de manera visual y en los apartados físicos, posteriormente realizar una comparativa entre los resultados y los datos de entrenamiento, se proponen los retos para usar estas herramientas en el estudio de la fotósfera solar, tubos de flujo y poros. (Texto tomado de la fuente).spa
dc.description.abstractThe present work is framed in the applications of deep neural networks for the modeling of the phenomena present in the solar photosphere. The proposed research is based on the construction of a 3D deep generative convolutional neural network, DCGAN, using Python artificial intelligence modules such as Pytorch for neural network architecture. It is intended to train a neural network capable of generating groups of cubes of high similarity with training cubes, these cubes correspond to physical quantities of the solar photosphere such as density, magnetic field, plasma velocity, temperature, among others, obtained from the simulation code MURaM. Simulation code developed by the Solar-MHD group of the Max Planck Institute developed between 2001-2005 with the purpose of generating realistic simulations of magneto-convection processes and magnetic activities, which have an effect on the convective zone of the sun, The present work seeks to take its results and take them as training data for the neural network generating new data with a similarity in a visual way and in the physical sections, then make a comparison between the results and the training data, the challenges are proposed to use these tools in the study of the solar photosphere, flux tubes and pores.eng
dc.description.degreelevelMaestríaspa
dc.description.degreenameMagíster en Ciencias - Astronomíaspa
dc.description.researchareaAstrofísica solarspa
dc.format.extentviii, 66 páginasspa
dc.format.mimetypeapplication/pdfspa
dc.identifier.instnameUniversidad Nacional de Colombiaspa
dc.identifier.reponameRepositorio Institucional Universidad Nacional de Colombiaspa
dc.identifier.repourlhttps://repositorio.unal.edu.co/spa
dc.identifier.urihttps://repositorio.unal.edu.co/handle/unal/85995
dc.language.isospaspa
dc.publisherUniversidad Nacional de Colombiaspa
dc.publisher.branchUniversidad Nacional de Colombia - Sede Bogotáspa
dc.publisher.facultyFacultad de Cienciasspa
dc.publisher.placeBogotá, Colombiaspa
dc.publisher.programBogotá - Ciencias - Maestría en Ciencias - Astronomíaspa
dc.relation.referencesM. Schüssler F. Cattaneo T. Emonet A. Vögler, S. Shelyag1 and T. Linde. Simulations of magneto- convection in the solar photosphere equations, methods, and results of the muram code. As- tronomy & Astrophysics, II:8–10, 2004.spa
dc.relation.referencesC. M. Bishop. Neural networks for pattern recognition. Astronomy & Astrophysics, II:0–498, 1996.spa
dc.relation.referencesRoger P Briggs. Solar Physics and Terrestrial Effects: A Curriculum Guide for Teachers, Grades 7–12. Space Environment Laboratory, National Oceanic and Atmospheric Administration, 1993.spa
dc.relation.referencesI. Dabbura. Gradient descent algorithm and its variants, December 21 2017. URL https://towardsdatascience.com/gradient-descent-algorithm-and-its-variants-10f652806a3. towards Data Science.spa
dc.relation.referencesLore Goetschalckx, Alex Andonian, and Johan Wagemans. Generative adversarial networks un- lock new methods for cognitive science. Trends in Cognitive Sciences, September 2021. doi: 10.1016/j.tics.2021.07.002.spa
dc.relation.referencesGeorge Ellery Hale. On the probable existence of a magnetic field in sun-spots. Astrophysical Journal, 28:315–343, 1908.spa
dc.relation.referencesRom Harré. Models in science. Physics Education, 1978. URL https://api.semanticscholar. org/CorpusID:117952040.spa
dc.relation.referencesYoshua Bengio Ian Goodfellow and Aaron Courville. Deep learning. The MIT Press, London, 2016.spa
dc.relation.referencesJames B Kaler. Stars and their spectra: an introduction to the spectral sequence. Cambridge University Press, 2011.spa
dc.relation.referencesSandesh Kumar. Redes neuronales de convolución en pocas palabras, julio 5 2020. URL https: //www.herevego.com/redes-neuronales-de-convolucion/. Home; Sandesh Kumar.spa
dc.relation.referencesJoan Manuel Bullon Lahuerta and Ma Ángeles del Castillo Alarcos. Observación solar. APEA, Aso- ciación para la Enseñanza de la Astronomía, 2010.spa
dc.relation.referencesMaud Langlois, Gil Moretto, Kit Richards, Steve Hegwer, and Thomas R Rimmele. Solar multi- conjugate adaptive optics at the dunn solar telescope: preliminary results. In Advancements in Adaptive Optics, volume 5490, pages 59–66. SPIE, 2004.spa
dc.relation.referencesMcCulloch, JY Pitts, Lettvin, PD Wall, and B Howland. Inhibition of impulses travelling in the primary afferent dorsal column fibres of cat spinal cord. In Fed Proc, volume 13, page 87, 1954.spa
dc.relation.referencesDale A Ostlie and Bradley W Carroll. An introduction to modern astrophysics. Addison-Wesley Read- ing, MA, USA, 2007.spa
dc.relation.referencesDina Prialnik. An Introduction to the Theory of Stellar Structure and Evolution. CAMBRIDGE, Cambridge, 2010.spa
dc.relation.referencesEric Ronald Priest. Solar magnetohydrodynamics, volume 21. Springer Science & Business Media, 2012.spa
dc.relation.referencesDavid E Rumelhart, Geoffrey E Hinton, and Ronald J Williams. Learning representations by back-propagating errors. nature, 323(6088):533–536, 1986.spa
dc.relation.referencesKarl Schwarzschild. Gesammelte Werke Collected Works: Volume 1, volume 1. Springer-Verlag, 2013.spa
dc.relation.referencesSergiy Shelyag, Peter Keys, Mihalis Mathioudakis, and Francis P Keenan. Vorticity in the solar photosphere. Astronomy & Astrophysics, 526:A5, 2011.spa
dc.relation.referencesK Wilhelm, P Lemaire, W Curdt, U Schühle, E Marsch, AI Poland, SD Jordan, RJ Thomas, DM Has- sler, MCE Huber, et al. First results of tide sumer telescope and spectrometer on soho: I. spectra and spectroradiometry. The First Results from SOHO, pages 75–104, 1997.spa
dc.relation.referencesWilliam Falcon et al. pytorch-lightning.readthedocs.io, Enero de 2024.spa
dc.relation.referencesFriedrich Wöger, Thomas Rimmele, Andrew Ferayorni, Andrew Beard, Brian S Gregory, Predrag Sekulic, and Steven L Hegwer. The daniel k. inouye solar telescope (dkist)/visible broadband imager (vbi). Solar Physics, 296:1–25, 2021.spa
dc.relation.referencesHarold Zirin. Astrophysics of the Sun. Cambridge University Press, Cambridge, UK, 1988. ISBN 978-0521343540.spa
dc.relation.referencesAbdurrahman Öcal and Lale Özbakır. Supervised deep convolutional generative adversarial net- works. Neurocomputing, 449:389–398, 2021. ISSN 0925-2312. doi: https://doi.org/10.1016/ j.neucom.2021.03.125. URL https://www.sciencedirect.com/science/article/pii/ S0925231221005178.spa
dc.relation.referencesW Dean Pesnell, B J. Thompson, and PC Chamberlin. The solar dynamics observatory (SDO). Springer, 2012.spa
dc.rights.accessrightsinfo:eu-repo/semantics/openAccessspa
dc.rights.licenseAtribución-NoComercial-SinDerivadas 4.0 Internacionalspa
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/4.0/spa
dc.subject.ddc520 - Astronomía y ciencias afines::522 - Técnicas, procedimientos, aparatos, equipos, materialesspa
dc.subject.proposalDCGANspa
dc.subject.proposalPytorcheng
dc.subject.proposalFotósferaspa
dc.subject.proposalMHDspa
dc.subject.proposalDeep learningspa
dc.subject.proposalCNNspa
dc.subject.proposalPhotosphereeng
dc.subject.wikidataAprendizaje profundospa
dc.subject.wikidatadeep learningeng
dc.subject.wikidataFotosferaspa
dc.subject.wikidataphotosphereeng
dc.subject.wikidataastronomía solarspa
dc.subject.wikidatasolar astronomyeng
dc.titleAplicación de técnicas de Deep Learning en modelamiento y observación de la fotósfera solarspa
dc.title.translatedApplication of Deep Learning techniques in modeling and observation of the photosphereeng
dc.typeTrabajo de grado - Maestríaspa
dc.type.coarhttp://purl.org/coar/resource_type/c_bdccspa
dc.type.coarversionhttp://purl.org/coar/version/c_ab4af688f83e57aaspa
dc.type.contentTextspa
dc.type.driverinfo:eu-repo/semantics/masterThesisspa
dc.type.redcolhttp://purl.org/redcol/resource_type/TMspa
dc.type.versioninfo:eu-repo/semantics/acceptedVersionspa
dcterms.audience.professionaldevelopmentEstudiantesspa
dcterms.audience.professionaldevelopmentInvestigadoresspa
oaire.accessrightshttp://purl.org/coar/access_right/c_abf2spa

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
1023010074.2024.pdf
Tamaño:
40.96 MB
Formato:
Adobe Portable Document Format
Descripción:
Tesis de Maestría en Ciencias - Astronomía

Bloque de licencias

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
license.txt
Tamaño:
5.74 KB
Formato:
Item-specific license agreed upon to submission
Descripción: