Ce travail présente un développement et une analyse des approches numériques déterministes et probabilistes efficaces pour les équations aux dérivées partielles avec des coefficients et données aléatoires. On s'intéresse au problème d'écoulement stationnaire avec des données aléatoires. Une méthode de projection dans le cas unidimensionnel est présentée, permettant de calculer efficacement la moyenne de la solution. Nous utilisons la méthode de collocation anisotrope des grilles clairsemées. D'abord, un indicateur de l'erreur satisfaisant une borne supérieure de l'erreur est introduit, il permet de calculer les poids d'anisotropie de la méthode. Ensuite, nous démontrons une amélioration de l'erreur a priori de la méthode. Elle confirme l'efficacité de la méthode en comparaison avec Monte-Carlo et elle sera utilisée pour accélérer la méthode par l'extrapolation de Richardson. Nous présentons aussi une analyse numérique d'une méthode probabiliste pour quantifier la migration d'un contaminant dans un milieu aléatoire. Nous considérons le problème d'écoulement couplé avec l'équation d'advection-diffusion, où on s'intéresse à la moyenne de l'extension et de la dispersion du soluté. Le modèle d'écoulement est discrétisée par une méthode des éléments finis mixtes, la concentration du soluté est une densité d'une solution d'une équation différentielle stochastique, qui sera discrétisée par un schéma d'Euler. Enfin, on présente une formule explicite de la dispersion et des estimations de l'erreur a priori optimales. / This work presents a development and an analysis of an effective deterministic and probabilistic approaches for partial differential equation with random coefficients and data. We are interesting in the steady flow equation with stochastic input data. A projection method in the one-dimensional case is presented to compute efficiently the average of the solution. An anisotropic sparse grid collocation method is also used to solve the flow problem. First, we introduce an indicator of the error satisfying an upper bound of the error, it allows us to compute the anisotropy weights of the method. We demonstrate an improvement of the error estimation of the method which confirms the efficiency of the method compared with Monte Carlo and will be used to accelerate the method using the Richardson extrapolation technique. We also present a numerical analysis of one probabilistic method to quantify the migration of a contaminant in random media. We consider the previous flow problem coupled with the advection-diffusion equation, where we are interested in the computation of the mean extension and the mean dispersion of the solute. The flow model is discretized by a mixed finite elements method and the concentration of the solute is a density of a solution of the stochastic differential equation, this latter will be discretized by an Euler scheme. We also present an explicit formula of the dispersion and an optimal a priori error estimates.

Ce travail présente quelques résultats concernant des méthodes numériques déterministes et probabilistes pour des équations aux dérivées partielles à coefficients aléatoires, avec des applications à l'hydrogéologie. On s'intéresse tout d'abord à l'équation d'écoulement dans un milieu poreux en régime stationnaire avec un coefficient de perméabilité lognormal homogène, incluant le cas d'une fonction de covariance peu régulière. On établit des estimations aux sens fort et faible de l'erreur commise sur la solution en tronquant le développement de Karhunen-Loève du coefficient. Puis on établit des estimations d'erreurs éléments finis dont on déduit une extension de l'estimation d'erreur existante pour la méthode de collocation stochastique, ainsi qu'une estimation d'erreur pour une méthode de Monte-Carlo multi-niveaux. On s'intéresse enfin au couplage de l'équation d'écoulement considérée précédemment avec une équation d'advection-diffusion, dans le cas d'incertitudes importantes et d'une faible longueur de corrélation. On propose l'analyse numérique d'une méthode numérique pour calculer la vitesse moyenne à laquelle la zone contaminée par un polluant s'étend. Il s'agit d'une méthode de Monte-Carlo combinant une méthode d'élements finis pour l'équation d'écoulement et un schéma d'Euler pour l'équation différentielle stochastique associée à l'équation d'advection-diffusion, vue comme une équation de Fokker-Planck. / This work presents some results about probabilistic and deterministic numerical methods for partial differential equations with stochastic coefficients, with applications to hydrogeology. We first consider the steady flow equation in porous media with a homogeneous lognormal permeability coefficient, including the case of a low regularity covariance function. We establish error estimates, both in strong and weak senses, of the error in the solution resulting from the truncature of the Karhunen-Loève expansion of the coefficient. Then we establish finite element error estimates, from which we deduce an extension of the existing error estimate for the stochastic collocation method along with an error estimate for a multilevel Monte-Carlo method. We finally consider the coupling of the previous flow equation with an advection-diffusion equation, in the case when the uncertainty is important and the correlation length is small. We propose the numerical analysis of a numerical method, which aims at computing the mean velocity of the expansion of a pollutant. The method consists in a Monte-Carlo method, combining a finite element method for the flow equation and an Euler scheme for the stochastic differential equation associated to the advection-diffusion equation, seen as a Fokker-Planck equation.

Les travaux de la présente thèse portent sur le recalage de modèles dynamiques d’aubes de turbinecomposites à matrice céramique. Ils s’inscrivent dans le cadre de la quantification d’incertitudes pour la validation de modèles et ont pour objectif de fournir des outils d’aide à la décision pour les ingénieurs desbureaux d’études. En effet, la dispersion importante observée lors des campagnes expérimentales invalidel’utilisation des méthodes de recalage déterministe. Après un état de l’art sur la relation entre les incertitudeset la physique, l’approche de Vérification & Validation a été introduite comme approche permettantd’assurer la crédibilité des modèles numériques. Puis, deux méthodes de recalages stochastiques, permettantde déterminer la distribution statistique des paramètres, ont été comparées sur un cas académique. La priseen compte des incertitudes n’élude pas les potentielles compensations entre paramètres. Par conséquent, desindicateurs ont été développés afin de détecter la présence de ces phénomènes perturbateurs. Ensuite, lathéorie info-gap a été employée en tant que moyen de modéliser ces méconnaissances. Une méthode derecalage stochastique robuste a ainsi été proposée, assurant un compromis entre la fidélité du modèle auxessais et la robustesse aux méconnaissances. Ces outils ont par la suite été appliqués sur un modèle éléments / This work is focused on the stochastic updating of ceramic matrix composite turbine blade model. They arepart of the uncertainty quantification framework for model validation. The aim is to enhance the existing toolused by the industrial decision makers. Indeed, consequent dispersion was measured during the experimentalcampaigns preventing the use of deterministic approaches. The first part of this thesis is dedicated to therelationship between mechanical science and uncertainty. Thus, Verification and Validation was introduced asthe processes by which credibility in numerical models is established. Then two stochastic updatingtechniques, able to handle statistic distribution, were compared through an academic example. Nevertheless,taking into account uncertainties doesn’t remove potential compensating effects between parameters.Therefore, criteria were developed in order to detect these disturbing phenomena. Info-gap theory wasemployed as a mean to model these lack of knowledge. Paired with the stochastic updating method, a robuststochasticapproach has been proposed. Results demonstrate a trade-off relationship between the model’sfidelity and robustness. The developed tools were applied on a ceramic matrix composite turbine blade finiteelement model.

Bayesian inference occurs when prior knowledge about uncertain parameters in mathematical models is merged with new observational data related to the model outcome. In this thesis we focus on models given by partial differential equations where the uncertain parameters are coefficient functions belonging to infinite dimensional function spaces. The result of the Bayesian inference is then a well-defined posterior probability measure on a function space describing the updated knowledge about the uncertain coefficient. For decision making and post-processing it is often required to sample or integrate wit resprect to the posterior measure. This calls for sampling or numerical methods which are suitable for infinite dimensional spaces. In this work we focus on Kalman filter techniques based on ensembles or polynomial chaos expansions as well as Markov chain Monte Carlo methods. We analyze the Kalman filters by proving convergence and discussing their applicability in the context of Bayesian inference. Moreover, we develop and study an improved dimension-independent Metropolis-Hastings algorithm. Here, we show geometric ergodicity of the new method by a spectral gap approach using a novel comparison result for spectral gaps. Besides that, we observe and further analyze the robustness of the proposed algorithm with respect to decreasing observational noise. This robustness is another desirable property of numerical methods for Bayesian inference. The work concludes with the application of the discussed methods to a real-world groundwater flow problem illustrating, in particular, the Bayesian approach for uncertainty quantification in practice. / Bayessche Inferenz besteht daraus, vorhandenes a-priori Wissen über unsichere Parameter in mathematischen Modellen mit neuen Beobachtungen messbarer Modellgrößen zusammenzuführen. In dieser Dissertation beschäftigen wir uns mit Modellen, die durch partielle Differentialgleichungen beschrieben sind. Die unbekannten Parameter sind dabei Koeffizientenfunktionen, die aus einem unendlich dimensionalen Funktionenraum kommen. Das Resultat der Bayesschen Inferenz ist dann eine wohldefinierte a-posteriori Wahrscheinlichkeitsverteilung auf diesem Funktionenraum, welche das aktualisierte Wissen über den unsicheren Koeffizienten beschreibt. Für Entscheidungsverfahren oder Postprocessing ist es oft notwendig die a-posteriori Verteilung zu simulieren oder bzgl. dieser zu integrieren. Dies verlangt nach numerischen Verfahren, welche sich zur Simulation in unendlich dimensionalen Räumen eignen. In dieser Arbeit betrachten wir Kalmanfiltertechniken, die auf Ensembles oder polynomiellen Chaosentwicklungen basieren, sowie Markowketten-Monte-Carlo-Methoden. Wir analysieren die erwähnte Kalmanfilter, indem wir deren Konvergenz zeigen und ihre Anwendbarkeit im Kontext Bayesscher Inferenz diskutieren. Weiterhin entwickeln und studieren wir einen verbesserten dimensionsunabhängigen Metropolis-Hastings-Algorithmus. Hierbei weisen wir geometrische Ergodizität mit Hilfe eines neuen Resultates zum Vergleich der Spektrallücken von Markowketten nach. Zusätzlich beobachten und analysieren wir die Robustheit der neuen Methode bzgl. eines fallenden Beobachtungsfehlers. Diese Robustheit ist eine weitere wünschenswerte Eigenschaft numerischer Methoden für Bayessche Inferenz. Den Abschluss der Arbeit bildet die Anwendung der diskutierten Methoden auf ein reales Grundwasserproblem, was insbesondere den Bayesschen Zugang zur Unsicherheitsquantifizierung in der Praxis illustriert.

Cette thèse concerne le développement, l'extension et l'application d'une formulation stochastique des équations de la mécanique des fluides introduite par Mémin (2014). La vitesse petite échelle, non-résolue, est modélisée au moyen d'un champ aléatoire décorrélé en temps. Cela modifie l'expression de la dérivée particulaire et donc les équations de la mécanique des fluides. Les modèles qui en découlent sont dénommés modèles sous incertitude de position. La thèse s'articulent autour de l'étude successive de modèles réduits, de versions stochastiques du transport et de l'advection à temps long d'un champ de traceur par une vitesse mal résolue. La POD est une méthode de réduction de dimension, pour EDP, rendue possible par l'utilisation d'observations. L'EDP régissant l'évolution de la vitesse du fluide est remplacée par un nombre fini d'EDOs couplées. Grâce à la modélisation sous incertitude de position et à de nouveaux estimateurs statistiques, nous avons dérivé et simulé des versions réduites, déterministe et aléatoire, de l'équation de Navier-Stokes. Après avoir obtenu des versions aléatoires de plusieurs modèles océaniques, nous avons montré numériquement que ces modèles permettaient de mieux prendre en compte les petites échelles des écoulements, tout en donnant accès à des estimés de bonne qualité des erreurs du modèle. Ils permettent par ailleurs de mieux rendre compte des évènements extrêmes, des bifurcations ainsi que des phénomènes physiques réalistes absents de certains modèles déterministes équivalents. Nous avons expliqué, démontré et quantifié mathématiquement l'apparition de petites échelles de traceur, lors de l'advection par une vitesse mal résolu. Cette quantification permet de fixer proprement des paramètres de la méthode d'advection Lagrangienne, de mieux le comprendre le phénomène de mélange et d'aider au paramétrage des simulations grande échelle en mécanique des fluides. / This thesis develops, analyzes and demonstrates several valuable applications of randomized fluid dynamics models referred to as under location uncertainty. The velocity is decomposed between large-scale components and random time-uncorrelated small-scale components. This assumption leads to a modification of the material derivative and hence of every fluid dynamics models. Through the thesis, the mixing induced by deterministic low-resolution flows is also investigated. We first applied that decomposition to reduced order models (ROM). The fluid velocity is expressed on a finite-dimensional basis and its evolution law is projected onto each of these modes. We derive two types of ROMs of Navier-Stokes equations. A deterministic LES-like model is able to stabilize ROMs and to better analyze the influence of the residual velocity on the resolved component. The random one additionally maintains the variability of stable modes and quantifies the model errors. We derive random versions of several geophysical models. We numerically study the transport under location uncertainty through a simplified one. A single realization of our model better retrieves the small-scale tracer structures than a deterministic simulation. Furthermore, a small ensemble of simulations accurately predicts and describes the extreme events, the bifurcations as well as the amplitude and the position of the ensemble errors. Another of our derived simplified model quantifies the frontolysis and the frontogenesis in the upper ocean. This thesis also studied the mixing of tracers generated by smooth fluid flows, after a finite time. We propose a simple model to describe the stretching as well as the spatial and spectral structures of advected tracers. With a toy flow but also with satellite images, we apply our model to locally and globally describe the mixing, specify the advection time and the filter width of the Lagrangian advection method, as well as the turbulent diffusivity in numerical simulations.

[ES] Las grandes presas, así como los diques de protección, son infraestructuras críticas cuyo fallo puede conllevar importantes consecuencias económicas y sociales. Tradicionalmente, la gestión del riesgo y la definición de estrategias de adaptación en la toma de decisiones han asumido la invariabilidad de las condiciones climáticas, incluida la persistencia de patrones históricos de variabilidad natural y la frecuencia de eventos extremos. Sin embargo, se espera que el cambio climático afecte de forma importante a los sistemas hídricos y comprometa la seguridad de las presas, lo que puede acarrear posibles impactos negativos en términos de costes económicos, sociales y ambientales. Los propietarios y operadores de presas deben por tanto adaptar sus estrategias de gestión y adaptación a medio y largo plazo a los nuevos escenarios climáticos. En la presente tesis se ha desarrollado una metodología integral para incorporar los impactos del cambio climático en la gestión de la seguridad de presas y en el apoyo a la toma de decisiones. El objetivo es plantear estrategias de adaptación que incorporen la variabilidad de los futuros riesgos, así como la incertidumbre asociada a los nuevos escenarios climáticos. El impacto del cambio climático en la seguridad de presas se ha estructurado utilizando modelos de riesgo y mediante una revisión bibliográfica interdisciplinaria sobre sus potenciales efectos. Esto ha permitido establecer un enfoque dependiente del tiempo que incorpore la evolución futura del riesgo, para lo cual se ha definido un nuevo indicador que evalúa cuantitativamente la eficiencia a largo plazo de las medidas de reducción de riesgo. Además, para integrar la incertidumbre de los escenarios futuros en la toma de decisiones, la metodología propone una estrategia robusta que permite establecer secuencias optimizadas de implementación de medidas correctoras para la adaptación al cambio climático. A pesar de las dificultades para asignar probabilidades a eventos específicos, esta metodología permite un análisis sistemático y objetivo, reduciendo considerablemente la subjetividad. Esta metodología se ha aplicado al caso real de una presa española susceptible a los efectos del cambio climático. El análisis se centra en el escenario hidrológico, donde las avenidas son la principal carga a la que está sometida la presa. Respecto de análisis previos de la presa, los resultados obtenidos proporcionan nueva y valiosa información sobre la evolución de los riesgos futuros y sobre cómo abordarlos. En general, se espera un aumento del riesgo con el tiempo; esto ha llevado a plantear nuevas medidas de adaptación que no están justificadas en la situación actual. Esta es la primera aplicación documentada de un análisis exhaustivo de los impactos del cambio climático sobre el riesgo de rotura de una presa que sirve como marco de referencia para la definición de estrategias de adaptación a largo plazo y la evaluación de su eficiencia. / [CAT] Les grans preses, així com els dics de protecció, són infraestructures crítiques que si fallen poden produir importants conseqüències econòmiques i socials. Tradicionalment, la gestió del risc i la definició d'estratègies d'adaptació en la presa de decisions han assumit la invariabilitat de les condicions climàtiques, inclosa la persistència de patrons històrics de variabilitat natural i la probabilitat d'esdeveniments extrems. No obstant això, s'espera que el canvi climàtic afecte de manera important als sistemes hídrics i comprometi la seguretat de les preses, la qual cosa pot implicar possibles impactes negatius en termes de costos econòmics, socials i ambientals. Els propietaris i operadors de preses deuen per tant adaptar les seues estratègies de gestió i adaptació a mitjà i llarg termini als nous escenaris climàtics. En la present tesi s'ha desenvolupat una metodologia integral per a incorporar els impactes del canvi climàtic en la gestió de la seguretat de preses i en el suport a la presa de decisions. L'objectiu és plantejar estratègies d'adaptació que incorporen la variabilitat dels futurs riscos, així com la incertesa associada als nous escenaris climàtics. L'impacte del canvi climàtic en la seguretat de preses s'ha estructurat utilitzant models de risc i mitjançant una revisió bibliogràfica interdisciplinària sobre els seus potencials efectes. Això ha permès establir un enfocament dependent del temps que incorpori l'evolució futura del risc, per a això s'ha definit un nou indicador que avalua quantitativament l'eficiència a llarg termini de les mesures de reducció de risc. A més, per a integrar la incertesa dels escenaris futurs en la presa de decisions, la metodologia proposa una estratègia robusta que permet establir seqüències optimitzades d'implementació de mesures correctores per a l'adaptació al canvi climàtic. A pesar de les dificultats per a assignar probabilitats a esdeveniments específics, esta metodologia permet una anàlisi sistemàtica i objectiva, reduint considerablement la subjectivitat. Aquesta metodologia s'ha aplicat al cas real d'una presa espanyola susceptible a l'efecte del canvi climàtic. L'anàlisi se centra en l'escenari hidrològic, on les avingudes són la principal càrrega a la qual està sotmesa la presa. Respecte d'anàlisis prèvies de la presa, els resultats obtinguts proporcionen nova i valuosa informació sobre l'evolució dels riscos futurs i sobre com abordar-los. En general, s'espera un augment del risc amb el temps; això ha portat a plantejar noves mesures d'adaptació que no estarien justificades en la situació actual. Aquesta és la primera aplicació documentada d'una anàlisi exhaustiva dels impactes del canvi climàtic sobre el risc de trencament d'una presa que serveix com a marc de referència per a la definició d'estratègies d'adaptació a llarg termini i l'avaluació de la seua eficiencia. / [EN] Large dams as well as protective dikes and levees are critical infrastructures whose failure has major economic and social consequences. Risk assessment approaches and decision-making strategies have traditionally assumed the stationarity of climatic conditions, including the persistence of historical patterns of natural variability and the likelihood of extreme events. However, climate change has a major impact on the world's water systems and is endangering dam safety, leading to potentially damaging impacts in terms of economic, social and environmental costs. Owners and operators of dams must adapt their mid- and long-term management and adaptation strategies to new climate scenarios. This thesis proposes a comprehensive approach to incorporate climate change impacts on dam safety management and decision-making support. The goal is to design adaptation strategies that incorporate the non-stationarity of future risks as well as the uncertainties associated with new climate scenarios. Based on an interdisciplinary review of the state-of-the-art research on its potential effects, the global impact of climate change on dam safety is structured using risk models. This allows a time-dependent approach to be established to consider the potential evolution of risk with time. Consequently, a new indicator is defined to support the quantitative assessment of the long-term efficiency of risk reduction measures. Additionally, in order to integrate the uncertainty of future scenarios, the approach is enhanced with a robust decision-making strategy that helps establish the consensus sequence of measures to be implemented for climate change adaptation. Despite the difficulties to allocate probabilities to specific events, such framework allows for a systematic and objective analysis, reducing considerably the subjectivity. Such a methodology is applied to a real case study of a Spanish dam subjected to the effects of climate change. The analysis focus on hydrological scenarios, where floods are the main load to which the dam is subjected. The results provide valuable new information with respect to the previously existing analysis of the dam regarding the evolution of future risks and how to cope with it. In general, risks are expected to increase with time and, as a result, new adaptation measures that are not justifiable for the present situation are recommended. This is the first documented application of a comprehensive analysis of climate change impacts on dam failure risk and serves as a reference benchmark for the definition of long-term adaptation strategies and the evaluation of their efficiency. / Fluixá Sanmartín, J. (2020). Adaptation strategies of dam safety management to new climate change scenarios informed by risk indicators [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/157634 / TESIS

Bayesian inference occurs when prior knowledge about uncertain parameters in mathematical models is merged with new observational data related to the model outcome. In this thesis we focus on models given by partial differential equations where the uncertain parameters are coefficient functions belonging to infinite dimensional function spaces. The result of the Bayesian inference is then a well-defined posterior probability measure on a function space describing the updated knowledge about the uncertain coefficient. For decision making and post-processing it is often required to sample or integrate wit resprect to the posterior measure. This calls for sampling or numerical methods which are suitable for infinite dimensional spaces. In this work we focus on Kalman filter techniques based on ensembles or polynomial chaos expansions as well as Markov chain Monte Carlo methods. We analyze the Kalman filters by proving convergence and discussing their applicability in the context of Bayesian inference. Moreover, we develop and study an improved dimension-independent Metropolis-Hastings algorithm. Here, we show geometric ergodicity of the new method by a spectral gap approach using a novel comparison result for spectral gaps. Besides that, we observe and further analyze the robustness of the proposed algorithm with respect to decreasing observational noise. This robustness is another desirable property of numerical methods for Bayesian inference. The work concludes with the application of the discussed methods to a real-world groundwater flow problem illustrating, in particular, the Bayesian approach for uncertainty quantification in practice. / Bayessche Inferenz besteht daraus, vorhandenes a-priori Wissen über unsichere Parameter in mathematischen Modellen mit neuen Beobachtungen messbarer Modellgrößen zusammenzuführen. In dieser Dissertation beschäftigen wir uns mit Modellen, die durch partielle Differentialgleichungen beschrieben sind. Die unbekannten Parameter sind dabei Koeffizientenfunktionen, die aus einem unendlich dimensionalen Funktionenraum kommen. Das Resultat der Bayesschen Inferenz ist dann eine wohldefinierte a-posteriori Wahrscheinlichkeitsverteilung auf diesem Funktionenraum, welche das aktualisierte Wissen über den unsicheren Koeffizienten beschreibt. Für Entscheidungsverfahren oder Postprocessing ist es oft notwendig die a-posteriori Verteilung zu simulieren oder bzgl. dieser zu integrieren. Dies verlangt nach numerischen Verfahren, welche sich zur Simulation in unendlich dimensionalen Räumen eignen. In dieser Arbeit betrachten wir Kalmanfiltertechniken, die auf Ensembles oder polynomiellen Chaosentwicklungen basieren, sowie Markowketten-Monte-Carlo-Methoden. Wir analysieren die erwähnte Kalmanfilter, indem wir deren Konvergenz zeigen und ihre Anwendbarkeit im Kontext Bayesscher Inferenz diskutieren. Weiterhin entwickeln und studieren wir einen verbesserten dimensionsunabhängigen Metropolis-Hastings-Algorithmus. Hierbei weisen wir geometrische Ergodizität mit Hilfe eines neuen Resultates zum Vergleich der Spektrallücken von Markowketten nach. Zusätzlich beobachten und analysieren wir die Robustheit der neuen Methode bzgl. eines fallenden Beobachtungsfehlers. Diese Robustheit ist eine weitere wünschenswerte Eigenschaft numerischer Methoden für Bayessche Inferenz. Den Abschluss der Arbeit bildet die Anwendung der diskutierten Methoden auf ein reales Grundwasserproblem, was insbesondere den Bayesschen Zugang zur Unsicherheitsquantifizierung in der Praxis illustriert.

Microelectromechanical systems (MEMS) depend on mechanical deformation to sense their environment, enhance electrical circuitry, or store data. Nonlinear forces arising from multiphysics phenomena at the micro- and nanoscale -- van der Waals forces, electrostatic fields, dielectric charging, capillary forces, surface roughness, asperity interactions -- lead to challenging problems for analysis, simulation, and measurement of the deforming device elements. Herein, a foundation for the study of mechanical deformation is provided through computational and experimental studies of MEMS microcantilever capacitive switches. Numerical techniques are built to capture deformation equilibria expediently. A compact analytical model is developed from principle multiphysics governing operation. Experimental measurements support the phenomena predicted by the analytical model, and finite element method (FEM) simulations confirm device-specific performance. Altogether, the static multistability and quasistatic performance of the electrostatically-actuated switches are confirmed across analysis, simulation, and experimentation. <p><br></p> <p>The nonlinear multiphysics forces present in the devices are critical to the switching behavior exploited for novel applications, but are also a culprit in a common failure mode when the attractive forces overcome the restorative and repulsive forces to result in two elements sticking together. Quasistatic operation is functional for switching between multistable states during normal conditions, but is insufficient under such stiction-failure. Exploration of dynamic methods for stiction release is often the only option for many system configurations. But how and when is release achieved? To investigate the fundamental mechanism of dynamic release, an atomic force microscopy (AFM) system -- a microcantilever with a motion-controlled base and a single-asperity probe tip, measured and actuated via lasers -- is configured to replicate elements of a stiction-failed MEMS device. Through this surrogate, observable dynamic signatures of microcantilever deflection indicate the onset of detachment between the probe and a sample.</p>

<p>The Ensemble Kalman filter (EnKF) has achieved great successes in data assimilation in atmospheric and oceanic sciences, but its failure in convergence to the right filtering distribution precludes its use for uncertainty quantification. Other existing methods, such as particle filter or sequential importance sampler, do not scale well to the dimension of the system and the sample size of the datasets. In this dissertation, we address these difficulties in a coherent way.</p><p><br></p><p> </p><p>In the first part of the dissertation, we reformulate the EnKF under the framework of Langevin dynamics, which leads to a new particle filtering algorithm, the so-called Langevinized EnKF (LEnKF). The LEnKF algorithm inherits the forecast-analysis procedure from the EnKF and the use of mini-batch data from the stochastic gradient Langevin-type algorithms, which make it scalable with respect to both the dimension and sample size. We prove that the LEnKF converges to the right filtering distribution in Wasserstein distance under the big data scenario that the dynamic system consists of a large number of stages and has a large number of samples observed at each stage, and thus it can be used for uncertainty quantification. We reformulate the Bayesian inverse problem as a dynamic state estimation problem based on the techniques of subsampling and Langevin diffusion process. We illustrate the performance of the LEnKF using a variety of examples, including the Lorenz-96 model, high-dimensional variable selection, Bayesian deep learning, and Long Short-Term Memory (LSTM) network learning with dynamic data.</p><p><br></p><p> </p><p>In the second part of the dissertation, we focus on two extensions of the LEnKF algorithm. Like the EnKF, the LEnKF algorithm was developed for Gaussian dynamic systems containing no unknown parameters. We propose the so-called stochastic approximation- LEnKF (SA-LEnKF) for simultaneously estimating the states and parameters of dynamic systems, where the parameters are estimated on the fly based on the state variables simulated by the LEnKF under the framework of stochastic approximation. Under mild conditions, we prove the consistency of resulting parameter estimator and the ergodicity of the SA-LEnKF. For non-Gaussian dynamic systems, we extend the LEnKF algorithm (Extended LEnKF) by introducing a latent Gaussian measurement variable to dynamic systems. Those two extensions inherit the scalability of the LEnKF algorithm with respect to the dimension and sample size. The numerical results indicate that they outperform other existing methods in both states/parameters estimation and uncertainty quantification.</p>

[pt] Um dos objetivos desta tese é analisar o custo computacional do método de Monte Carlo aplicado a um problema modelo de dinâmica, considerando incertezas na força de atrito. O sistema mecânico a ser estudado é composto por um oscilador de um grau de liberdade que se desloca sobre uma esteira em movimento. Considera-se a existência de atrito seco entre a massa do oscilador e a esteira. Devido a uma descontinuidade na força de atrito, a dinâmica resultante pode ser dividida em duas fases que se alternam, chamadas de stick e slip. Neste estudo, um parâmetro da força de atrito dinâmica é modelado como uma variável aleatória. A propagação de incerteza é estudada por meio da aplicação do método de Monte Carlo, considerando três abordagens diferentes para calcular aproximações da resposta dos problemas de valor inicial que modelam a dinâmica do problema: NV) aproximações numéricas calculadas usando método de Runge-Kutta de quarta e quinta ordens com passo de integração variável; NF) aproximações numéricas calculadas usando método de Runge-Kutta de quarta ordem com passo de integração fixo; AN) aproximação analítica obtida com o método de múltiplas escalas. Nas abordagens NV e NF, para cada valor de parâmetro, uma aproximação numérica foi calculada. Já para a AN, apenas uma aproximação analítica foi calculada e avaliada para os diferentes valores usados. Entre as variáveis aleatórias de interesse associadas ao custo computacional do método de Monte Carlo, encontram-se o tempo de execução e o espaço em disco consumido. Devido à propagação de incertezas, a resposta do sistema é um processo estocástico com uma sequência aleatória de fases de stick e slip. Essa sequência pode ser caracterizada pelas seguintes variáveis aleatórias: instantes de transição entre as fases de stick e slip, suas durações e o número de fases. Para estudar as variáveis associadas ao custo computacional e ao processo estocástico foram construídos modelos estatísticos, histogramas normalizados e gráficos de dispersão. O objetivo é estudar a dependência entre as variáveis do processo estocástico e o custo computacional. Porém, a construção destas análises não é simples devido à dimensão do problema e à impossibilidade de visualização das distribuições conjuntas de vetores aleatórios de três ou mais dimensões. / [en] One of the objectives of this thesis is to analyze the computational cost of the Monte Carlo method applied to a toy problem concerning the dynamics of a mechanical system with uncertainties in the friction force. The system is composed by an oscillator placed over a moving belt. The existence of dry friction between the two elements in contact is considered. Due to a discontinuity in the frictional force, the resulting dynamics can be divided into two alternating phases, called stick and slip. In this study, a parameter of the dynamic friction force is modeled as a random variable. Uncertainty propagation is analyzed by applying the Monte Carlo method, considering three different strategies to compute approximations to the initial value problems that model the system s dynamics: NV) numerical approximations computed with the Runge-Kutta method of 4th and 5th orders, with variable integration time-step; NF) numerical approximations computed with the Runge-Kutta method of 4th order, with a fixed integration time-step; AN) analytical approximation obtained with the multiple scale method. In the NV and NF strategies, for each parameter value, a numerical approximation was calculated, whereas for the AN strategy, only one analytical approximation was calculated and evaluated for the different values of parameters considered. The run-time and the storage are among the random variables of interest associated with the computational cost of the Monte Carlo method. Due to uncertainty propagation, the system response is a stochastic process given by a random sequence of stick and slip phases. This sequence can be characterized by the following random variables: the transition instants between the stick and slip phases, their durations and the number of phases. To study the random processes and the variables related to the computational costs, statistical models, normalized histograms and scatterplots were built. Afterwards, a joint analysis was performed to study the dependece between the variables of the random process and the computational cost. However, the construction of these analyses is not a simple task due to the impossibility of viewing the distributionto of joint distributions of random vectors of three or more.