Os progressos computacionais nas últimas décadas e a teoria matemática cada vez mais sólida têm possibilitado a resolução de problemas de alta complexidade, permitindo uma modelagem cada vez mais detalhada da realidade. Tal verdade aplica-se inclusive para os sistemas rígidos de Equações Diferencias Ordinárias (EDOs): existem métodos numéricos altamente performáticos para este tipo de problema, que permitem uma grande variação no tamanho do passo de integração sem impactar na sua convergência. Este trabalho apresenta um estudo sobre o conceito de rigidez e técnicas numéricas para resolução de problemas rígidos de EDOs. O que nos motivou a estudar tais técnicas foram problemas oriundos da Física Nuclear que envolvem cadeias de decaimento radioativo. Estes problemas podem ser modelados por uma cadeia fechada de compartimentos que se traduz em um sistema de EDOs. Os elementos destas cadeias podem possuir constantes de decaimento com ordens de grandeza muito distintas, caracterizando a sua rigidez e exigindo cautela na resolução das equações que as modelam. Embora seja possível determinar a solução analítica para estes problemas, o uso de métodos numéricos facilita a obtenção da solução quando consideramos sistemas com um número elevado de equações. Além disso, soluções numéricas permitem adaptações na modelagem ou em ajustes de dados com mais facilidade. Métodos implícitos são indicados para a resolução deste tipo de problema, pois possuem uma região de estabilidade ilimitada. Neste trabalho, implementamos dois métodos numéricos que possuem esta característica: o método de Radau II e o método de Rosenbrock. Estes métodos foram utilizados para obtenção de soluções numéricas robustas para problemas rígidos de decaimento radioativo envolvendo cadeias naturais e artificiais, considerando retiradas de elementos das cadeias durante o processo de decaimento e quando queremos determinar qual era o estado inicial de uma cadeia que está em decaimento. Ambos os métodos foram implementados com estratégias de controle do tamanho do passo de integração e produziram resultados consistentes dentro de uma precisão pré-fixada. / The computational progress in the last decades and the increasingly solid mathematical theory have made possible the resolution of highly complex problems allowing an ever more detailed modelling of reality. This is true even for the systems of stiff Ordinary Differential Equations (ODEs): there are highly performative numerical methods for this kind of problem which allow a wide variation in the size of integration step without impacting on their convergence. This thesis presents a study about the concept of stiffness and numerical techniques to solve stiff problems of ODEs. What motivated us to study these techniques were problems from the Nuclear Physics involving radioactive decay chains. These problems could be modelled by a closed chain of compartments which is translated into a system of ODEs. The elements of these chains could have decay constants with very different orders of magnitude which characterizes the stiffness of the problem and requires caution in solving the model equations. Although it is possible to determine the analytical solution to these problems when we consider systems with a high number of equations, calculate the solution by numerical methods becomes easier. Furthermore, numerical solutions allow adaptations in modelling or data adjustments more easily. Implicit methods are indicated to solve this kind of problem because they have an unlimited region of stability. In this study, we implemented two numerical methods which have this feature: Radau II method and Rosenbrock method. These methods were used to obtain robust numerical solutions for stiff problems of radioactive decay involving natural and artificial chains, considering the removal of elements during the decay process and when we want to determine what was the initial state of a chain which is decaying. Both methods were implemented with control strategies for integration step size providing consistent results within a pre-established accuracy.

Os progressos computacionais nas últimas décadas e a teoria matemática cada vez mais sólida têm possibilitado a resolução de problemas de alta complexidade, permitindo uma modelagem cada vez mais detalhada da realidade. Tal verdade aplica-se inclusive para os sistemas rígidos de Equações Diferencias Ordinárias (EDOs): existem métodos numéricos altamente performáticos para este tipo de problema, que permitem uma grande variação no tamanho do passo de integração sem impactar na sua convergência. Este trabalho apresenta um estudo sobre o conceito de rigidez e técnicas numéricas para resolução de problemas rígidos de EDOs. O que nos motivou a estudar tais técnicas foram problemas oriundos da Física Nuclear que envolvem cadeias de decaimento radioativo. Estes problemas podem ser modelados por uma cadeia fechada de compartimentos que se traduz em um sistema de EDOs. Os elementos destas cadeias podem possuir constantes de decaimento com ordens de grandeza muito distintas, caracterizando a sua rigidez e exigindo cautela na resolução das equações que as modelam. Embora seja possível determinar a solução analítica para estes problemas, o uso de métodos numéricos facilita a obtenção da solução quando consideramos sistemas com um número elevado de equações. Além disso, soluções numéricas permitem adaptações na modelagem ou em ajustes de dados com mais facilidade. Métodos implícitos são indicados para a resolução deste tipo de problema, pois possuem uma região de estabilidade ilimitada. Neste trabalho, implementamos dois métodos numéricos que possuem esta característica: o método de Radau II e o método de Rosenbrock. Estes métodos foram utilizados para obtenção de soluções numéricas robustas para problemas rígidos de decaimento radioativo envolvendo cadeias naturais e artificiais, considerando retiradas de elementos das cadeias durante o processo de decaimento e quando queremos determinar qual era o estado inicial de uma cadeia que está em decaimento. Ambos os métodos foram implementados com estratégias de controle do tamanho do passo de integração e produziram resultados consistentes dentro de uma precisão pré-fixada. / The computational progress in the last decades and the increasingly solid mathematical theory have made possible the resolution of highly complex problems allowing an ever more detailed modelling of reality. This is true even for the systems of stiff Ordinary Differential Equations (ODEs): there are highly performative numerical methods for this kind of problem which allow a wide variation in the size of integration step without impacting on their convergence. This thesis presents a study about the concept of stiffness and numerical techniques to solve stiff problems of ODEs. What motivated us to study these techniques were problems from the Nuclear Physics involving radioactive decay chains. These problems could be modelled by a closed chain of compartments which is translated into a system of ODEs. The elements of these chains could have decay constants with very different orders of magnitude which characterizes the stiffness of the problem and requires caution in solving the model equations. Although it is possible to determine the analytical solution to these problems when we consider systems with a high number of equations, calculate the solution by numerical methods becomes easier. Furthermore, numerical solutions allow adaptations in modelling or data adjustments more easily. Implicit methods are indicated to solve this kind of problem because they have an unlimited region of stability. In this study, we implemented two numerical methods which have this feature: Radau II method and Rosenbrock method. These methods were used to obtain robust numerical solutions for stiff problems of radioactive decay involving natural and artificial chains, considering the removal of elements during the decay process and when we want to determine what was the initial state of a chain which is decaying. Both methods were implemented with control strategies for integration step size providing consistent results within a pre-established accuracy.

<p>  </p> <p>A <em>digital twin</em> (DT), which keeps track of nuclear reactor history to provide real-time predictions, has been recently proposed for nuclear reactor monitoring. A digital twin can be implemented using either a differential equations-based physics model, or a data-driven machine learning model<strong>. </strong>The principal challenge in physics model-based DT consists of achieving sufficient model fidelity to represent a complex experimental system, while the main challenge in data-driven DT appears in the extensive training requirements and potential lack of predictive ability. </p> <p>In this thesis, we investigate the performance of a hybrid approach, which is based on physics-informed neural networks (PINNs) that encode fundamental physical laws into the loss function of the neural network. In this way, PINNs establish theoretical constraints and biases to supplement measurement data and provide solution to several limitations of purely data-driven machine learning (ML) models. We develop a PINN model to solve the point kinetic equations (PKEs), which are time dependent stiff nonlinear ordinary differential equations that constitute a nuclear reactor reduced-order model under the approximation of ignoring the spatial dependence of the neutron flux. PKEs portray the kinetic behavior of the system, and this kind of approach is the basis for most analyses of reactor systems, except in cases where flux shapes are known to vary with time. This system describes the nuclear parameters such as neutron density concentration, the delayed neutron precursor density concentration and reactivity. Both neutron density and delayed neutron precursor density concentrations are the vital parameters for safety and the transient behavior of the reactor power. </p> <p>The PINN model solution of PKEs is developed to monitor a start-up transient of the Purdue University Reactor Number One (PUR-1) using experimental parameters for the reactivity feedback schedule and the neutron source. The facility under modeling, PUR-1, is a pool type small research reactor located in West Lafayette Indiana. It is an all-digital light water reactor (LWR) submerged into a deep-water pool and has a power output of 10kW. The results demonstrate strong agreement between the PINN solution and finite difference numerical solution of PKEs. We investigate PINNs performance in both data interpolation and extrapolation. </p> <p>The findings of this thesis research indicate that the PINN model achieved highest performance and lowest errors in data interpolation. In the case of extrapolation data, three different test cases were considered, the first where the extrapolation is performed in a five-seconds interval, the second where the extrapolation is performed in a 10-seconds interval, and the third where the extrapolation is performed in a 15-seconds interval. The extrapolation errors are comparable to those of interpolation predictions. Extrapolation accuracy decreases with increasing time interval.</p>