Ce travail porte sur la modélisation et la simulation du trafic routier sur un réseau. Modéliser le trafic sur une section homogène (c'est-à-dire sans entrée, ni sortie) trouve ses racines au milieu du XXème siècle et a généré une importante littérature depuis. Cependant, la prise en compte des discontinuités des réseaux comme les jonctions, n'a attiré l'attention du cercle scientifique que bien plus récemment. Pourtant, ces discontinuités sont les sources majeures des congestions, récurrentes ou non, qui dégradent la qualité de service des infrastructures. Ce travail se propose donc d'apporter un éclairage particulier sur cette question, tout en s'intéressant aux problèmes d'échelle et plus particulièrement au passage microscopique-macroscopique dans les modèles existants. La première partie de cette thèse est consacrée au lien existant entre les modèles de poursuite microscopiques et les modèles d'écoulement macroscopiques. Le passage asymptotique est assuré par une technique d'homogénéisation pour les équations d'Hamilton-Jacobi. Dans une deuxième partie, nous nous intéressons à la modélisation et à la simulation des flux de véhicules au travers d'une jonction. Le modèle macroscopique considéré est bâti autour des équations d'Hamilton-Jacobi. La troisième partie enfin, se concentre sur la recherche de solutions analytiques ou semi-analytiques, grâce à l'utilisation de formules de représentation permettant de résoudre les équations d'Hamilton-Jacobi sous de bonnes hypothèses. Nous nous intéressons également dans cette thèse, à la classe générique des modèles macroscopiques de trafic de second ordre, dits modèles GSOM / This work focuses on modeling and simulation of traffic flows on a network. Modeling road traffic on a homogeneous section takes its roots in the middle of XXth century and it has generated a substantial literature since then. However, taking into account discontinuities of the network such as junctions, has attracted the attention of the scientific circle more recently. However, these discontinuities are the major sources of traffic congestion, recurring or not, that basically degrades the level of service of road infrastructure. This work therefore aims to provide a unique perspective on this issue, while focusing on scale problems and more precisely on microscopic-macroscopic passage in existing models. The first part of this thesis is devoted to the relationship between microscopic car-following models and macroscopic continuous flow models. The asymptotic passage is based on a homogenization technique for Hamilton-Jacobi equations. In a second part, we focus on the modeling and simulation of vehicular traffic flow through a junction. The considered macroscopic model is built on Hamilton-Jacobi equations as well. Finally, the third part focuses on finding analytical or semi-analytical solutions, through representation formulas aiming to solve Hamilton-Jacobi equations under adequate assumptions. In this thesis, we are also interested in a generic class of second order macroscopic traffic flow models, the so-called GSOM models

Orientador: Edmundo Capelas de Oliveira / Dissertação (mestrado profissional) - Universidade Estadual de Campinas, Instituto de Matemática, Estatística e Computação Científica / Made available in DSpace on 2018-08-17T14:35:07Z (GMT). No. of bitstreams: 1 Rosa_EsterCristinaFontesdeAquino_M.pdf: 847998 bytes, checksum: d177526572b19cc1fdd5eeccdf511380 (MD5) Previous issue date: 2011 / Resumo: Este trabalho tem por objetivo modelar e resolver, matematicamente, um problema físico conhecido como pêndulo simples. Discutimos, como caso particular, as chamadas oscilações de pequena amplitude, isto é, uma aproximação que nos leva a mostrar que o período de oscilação é proporcional à raiz quadrada do quociente entre o comprimento do pêndulo e a aceleração da gravidade. Como vários outros problemas oriundos da Física, o pêndulo simples é representado através de equações diferenciais parciais. Assim, na busca de sua solução, aplicamos a metodologia de separação de variáveis que nos leva a um conjunto de equações ordinárias passíveis de simples integração. Escolhendo um sistema de coordenadas adequado, é conveniente usar o método de Hamilton-Jacobi, discutindo, antes, o problema do oscilador harmónico, apresentando, em seguida, o problema do pêndulo simples e impondo condições a fim de mostrar que as equações diferenciais associadas a esses dois sistemas são iguais, ou seja, suas soluções são equivalentes. Para tanto, estudamos o método de separação de variáveis associado às equações diferenciais parciais, lineares e de segunda ordem, com coeficientes constantes e três variáveis independentes, bem como a respectiva classificação quanto ao tipo. Posteriormente, estudamos as equações hipergeométricas, cujas soluções, as funções hipergeométricas. podem ser encontradas pelo método de Frobenius. Apresentamos o método de Hamilton-Jacobi, já mencionado, para o enfren-tamento do problema apresentado. Fizemos no capítulo final um apêndice sobre a função gama por sua presente importância no trato de funções hipergeométricas, em especial a integral elíptica completa de primeiro tipo que compõe a solução exata do período do pêndulo simples / Abstract: This work aims to present and solve, mathematically, the physics problem that is called simple pendulum. We reasoned, as an specific case, the so called low amplitude oscillation, that is, a convenient approximation that make us show that the period of oscillation is proportional to the quotient square root between the pendulum length and the gravity acceleration. Like several other problems arising from the physics, we are going to broach it through partial differential equations. Thus, in the search of its solution, we made use of the variable separation methodology that leads us to a body of ordinary equations susceptible of simple integration. Choosing an appropriate coordinate system, it is convenient to use the method Hamilton-Jacobi, arguing, first, the problem of the harmonic oscillator, with, then the problem of sf simple pendulum and imposing conditions to show that the differential equations associated with these two systems are equal, that is, their solutions are equivalent. With the purpose of reaching the objectives, we studied the variable separation method associated with partial differential equations, linear and of second order, with constant coefficient and three independent variables, as well as the respective classification about the type. Afterwards, we studied the hypergeometrical equations whose solutions, the hypergeometrical functions, are found by the Frobenius method. Introducing the Hamilton-Jacobi method, already mentioned, for addressing the problem presented. We made an appendix in the final chapter on the gamma function by its present importance in dealing with hypergeometric functions, in particular the elliptic integral of first kind consists of the exact period of sf simple pendulum / Mestrado / Fisica-Matematica / Mestre em Matemática

Inthis degree project, a solution on a coarse grid is recovered by fitting apartial differential equation to a few known data points. The PDE to consideris the heat equation and the Dupire’s equation with their synthetic data,including synthetic data from the Black-Scholes formula. The approach to fit aPDE is by optimal control to derive discrete approximations to regularized Hamiltoncharacteristic equations to which discrete stepping schemes, and parameters forsmoothness, are examined. By non-parametric numerical implementation thedervied method is tested and then a few suggestions on possible improvementsare given / I detta examensarbete återskapas en lösning på ett glest rutnät genom att anpassa en partiell differentialekvation till några givna datapunkter. De partiella differentialekvationer med deras motsvarande syntetiska data som betraktas är värmeledningsekvationen och Dupires ekvation inklusive syntetiska data från Black-Scholes formel. Tillvägagångssättet att anpassa en PDE är att med hjälp av optimal styrning härleda diskreta approximationer på ett system av regulariserade Hamilton karakteristiska ekvationer till vilka olika diskreta stegmetoder och parametrar för släthet undersöks. Med en icke-parametrisk numerisk implementation prövas den härledda metoden och slutligen föreslås möjliga förbättringar till metoden.

Cette thèse contient deux contributions à l’homogénéisation en espace-temps des équations de Hamilton-Jacobi du premier ordre. Ces équations sont en lien avec la modélisation du trafic routier. Enfin, sont présentés des résultats d’homogénéisation en milieu presque périodique. Le premier chapitre est consacré à l’homogénéisation d’un système infini d’équations différentielles couplées avec temps de retard. Ce système provient ici d’un modèle microscopique de trafic routier simple. Les conducteurs se suivent sur une route rectiligne infinie et l’on tient compte de leur temps de réaction. On suppose que la vitesse de chaque conducteur est une fonction de l’interdistance avec le conducteur qui le précède: on parle d’un modèle du type “follow-the-leader”. Grâce à un principe de comparaison strict, on montre la convergence vers un modèle macroscopique pour des temps de réaction inférieurs à une valeur critique. Dans un second temps, on exhibe un contre-exemple à l’homogénéisation pour un temps de réaction supérieur à cette valeur critique, pour des conditions initiales particulières. Pour cela, on perturbe la solution stationnaire dans laquelle les véhicules sont tous équidistants aux instants initiaux. Le second chapitre porte sur l’homogénéisation d’une équation de Hamilton-Jacobi dont l’Hamiltonien est discontinu en espace. Le modèle de trafic associé est une route rectiligne comportant une infinité de feux tricolores. Ces feux sont supposés identiques, équidistants et le déphasage entre deux feux successifs est supposé constant. On étudie l’influence à grande échelle de ce déphasage sur le trafic. On distingue la portion de route libre, qui sera représentée par un modèle macroscopique, et les feux, qui seront modélisés par des limiteurs de flux périodiques en temps. Le cadre théorique est celui par C. Imbert et R. Monneau (2017) pour les équations de Hamilton-Jacobi sur réseaux. L’étude se décompose en l’homogénéisation théorique, où l’Hamiltonien effectif dépend du déphasage, puis l’obtention de propriétés qualitatives de cet Hamiltonien à l’aide d’observations via des simulations numériques. Le troisième chapitre présente des résultats d’homogénéisation en milieu presque périodique. On étudie tout d’abord un problème d’évolution avec un Hamiltonien stationnaire, presque périodique en espace. À l’aide d’arguments presque périodiques, on effectue dans un second temps une nouvelle preuve du résultat d’homogénéisation du second chapitre. L’Hamiltonien est alors périodique en temps et presque périodique en espace. Sont également présentes des questions encore ouvertes, notamment dans le cas où l’Hamiltonien est presque périodique en temps-espace, et dans le cas d’un modèle de trafic où les feux sont assez proches, avec donc un modèle microscopique entre les feux / This thesis report deals with the homogenization in space and time of some first order Hamilton-Jacobi equations. It contains two contributions. The corresponding equations are derived from traffic flow modelling. We finally present some results of almost periodic homogenization. In the first chapter, we consider a one dimensional pursuit law with delay which is derived from traffic flow modelling. It takes the form of an infinite system of first order coupled delayed equations. Each equation describes the motion of a driver who interacts with the preceding one: such a model is referred to as a ``follow-the-leader" model. We take into account the reaction time of drivers. We derive a macroscopic model, namely a Hamilton-Jacobi equation, by a homogenization process for reaction times that are below an explicit threshold. The key idea is to show, that below this threshold, a strict comparison principle holds for the infinite system. Above this threshold, we show that collisions can occur. In a second time, for well-chosen dynamics and higher reaction times, we show that there exist some microscopic pursuit laws that do not lead to the previous macroscopic model. Such a law is here derived as a perturbation of the stationnary solution, for which all the vehicles are equally spaced at initial times. The second chapter is dedicated to the homogenization of a Hamilton-Jacobi equation for traffic lights. We consider an infinite road where lights are equally spaced and with a constant phase shift between two lights. This model takes the form of a first order Hamilton-Jacobi equation with an Hamiltonian that is discontinuous in the space variable and the notion of viscosity solution is the one introduced by C. Imbert and R. Monneau (2017). Each light is modelled as a time-periodic flux limiter and the traffic flow between two lights corresponds to the classical LWR model. The global Hamiltonian will be time-periodic but not periodic in space for a general phase shift. We first show that the rescaled solution converges toward the solution of the expected macroscopic model where the effective Hamiltonian depends on the phase shift. In a second time, numerical simulations are used to analyse the effect of the phase shift on the effective Hamiltonian and to reveal some properties of the effective Hamiltonian from the numerical observations. In the third chapter, we are interested in some homogenization problems of Hamilton-Jacobi equations within the almost periodic setting which generalizes the usual periodic one. The first problem is the evolutionary version of the work cite {ishii2000almost}, with the same stationary Hamiltonian. The second problem has already been solved in the second chapter but we use here almost periodic arguments for the time periodic and space almost periodic Hamiltonian. We only study the ergodicity of the associated cell problems. We finally discuss open problems, the first one concerning a space and time almost periodic Hamiltonian and the second one being a microscopic model for traffic flow modelling where the Hamiltonian is almost periodic in space

Cette thèse porte sur l'étude d'équation de Hamilton-Jacobi-Bellman associées à des problèmes de contrôle optimal et de jeux à champ moyen avec la particularité qu'on se place sur un réseau (c'est-à-dire, des ensembles constitués d'arêtes connectées par des jonctions) dans les deux problèmes, pour lesquels on autorise différentes dynamiques et différents coûts dans chaque bord d'un réseau. Dans la première partie de cette thèse, on considère un problème de contrôle optimal sur les réseaux dans l'esprit des travaux d'Achdou, Camilli, Cutrì & Tchou (2013) et Imbert, Moneau & Zidani (2013). La principale nouveauté est qu'on rajoute des coûts d'entrée (ou de sortie) aux sommets du réseau conduisant à une éventuelle discontinuité de la fonction valeur. Celle-ci est caractérisée comme l'unique solution de viscosité d'une équation Hamilton-Jacobi pour laquelle une condition de jonction adéquate est établie. L'unicité est une conséquence d'un principe de comparaison pour lequel nous donnons deux preuves différentes, l'une avec des arguments tirés de la théorie du contrôle optimal, inspirée par Achdou, Oudet & Tchou (2015) et l'autre basée sur les équations aux dérivées partielles, d'après Lions & Souganidis (2017). La deuxième partie concerne les jeux à champ moyen stochastiques sur les réseaux. Dans le cas ergodique, ils sont décrits par un système couplant une équation de Hamilton-Jacobi-Bellman et une équation de Fokker- Planck, dont les inconnues sont la densité m de la mesure invariante qui représente la distribution des joueurs, la fonction valeur v qui provient d'un problème de contrôle optimal "moyen" et la constante ergodique ρ. La fonction valeur v est continue et satisfait dans notre problème des conditions de Kirchhoff aux sommets très générales. La fonction m satisfait deux conditions de transmission aux sommets. En particulier, due à la généralité des conditions de Kirchhoff, m est en général discontinue aux sommets. L'existence et l'unicité d'une solution faible sont prouvées pour des Hamiltoniens sous-quadratiques et des hypothèses très générales sur le couplage. Enfin, dans la dernière partie, nous étudions les jeux à champ moyen stochastiques non stationnaires sur les réseaux. Les conditions de transition pour la fonction de valeur v et la densité m sont similaires à celles données dans la deuxième partie. Là aussi, nous prouvons l'existence et l'unicité d'une solution faible pour des Hamiltoniens sous-linéaires et des couplages et dans le cas d'un couplage non-local régularisant et borné inférieurement. La principale difficulté supplémentaire par rapport au cas stationnaire, qui nous impose des hypothèses plus restrictives, est d'établir la régularité des solutions du système posé sur un réseau. Notre approche consiste à étudier la solution de l'équation de Hamilton-Jacobi dérivée pour gagner de la régularité sur la solution de l'équation initiale. / The dissertation focuses on the study of Hamilton-Jacobi-Bellman equations associated with optimal control problems and mean field games problems in the case when the state space is a network. Different dynamics and running costs are allowed in each edge of the network. In the first part of this thesis, we consider an optimal control on networks in the spirit of the works of Achdou, Camilli, Cutrì & Tchou (2013) and Imbert, Monneau & Zidani (2013). The main new feature is that there are entry (or exit) costs at the edges of the network leading to a possible discontinuous value function. The value function is characterized as the unique viscosity solution of a Hamilton-Jacobi equation for which an adequate junction condition is established. The uniqueness is a consequence of a comparison principle for which we give two different proofs. One uses some arguments from the theory of optimal control and is inspired by Achdou, Oudet & Tchou (2015). The other one is based on partial differential equations techniques and is inspired by a recent work of Lions & Souganidis (2017). The second part is about stochastic mean field games for which the state space is a network. In the ergodic case, they are described by a system coupling a Hamilton- Jacobi-Bellman equation and a Fokker-Planck equation, whose unknowns are the density m of the invariant measure which represents the distribution of the players, the value function v which comes from an "average" optimal control problem and the ergodic constant ρ. The function v is continuous and satisfies general Kirchhoff conditions at the vertices. The density m satisfies dual transmission conditions. In particular, due to the generality of Kirchhoff’s conditions, m is in general discontinuous at the vertices. Existence and uniqueness are proven for subquadratic Hamiltonian and very general assumptions about the coupling term. Finally, in the last part, we study non-stationary stochastic mean field games on networks. The transition conditions for value function v and the density m are similar to the ones given in second part. Here again, we prove the existence and uniqueness of a weak solution for sublinear Hamiltonian and bounded non-local regularizing coupling term. The main additional difficulty compared to the stationary case, which imposes us more restrictive hypotheses, is to establish the regularity of the solutions of the system placed on a network. Our approach is to study the solution of the derived Hamilton-Jacobi equation to gain regularity over the initial equation.

Cette thèse porte sur l'étude de problèmes de contrôle optimal sur des réseaux (c'est-à-dire des ensembles constitués de sous-régions reliées entre elles par des jonctions), pour lesquels on autorise différentes dynamiques et différents coûts instantanés dans chaque sous-région du réseau. Comme dans les cas plus classiques, on aimerait pouvoir caractériser la fonction valeur d'un tel problème de contrôle par le biais d'une équation de Hamilton-Jacobi-Bellman. Cependant, les singularités géométriques du domaine, ainsi que les discontinuités des données ne nous permettent pas d'appliquer la théorie classique des solutions de viscosité. Dans la première partie de cette thèse nous prouvons que les fonctions valeurs de problèmes de contrôle optimal définis sur des réseaux 1-dimensionnel sont caractérisées par de telles équations. Dans la seconde partie les résultats précédents sont étendus au cas de problèmes de contrôle définis sur une jonction 2-dimensionnelle. Enfin, dans une dernière partie, nous utilisons les résultats obtenus précédemment pour traiter un problème de perturbation singulière impliquant des problèmes de contrôle optimal dans le plan pour lesquels les dynamiques et les coûts instantanés peuvent être discontinus à travers une frontière oscillante. / This thesis focuses on the study of optimal control problems defined on networks (i.e. sets consisting of sub-regions connected together through junctions), where different dynamics and different running costs are allowed in each sub-region of the network. As in classical cases, we would like to characterize the value function of such an optimal control problem through an Hamilton-Jacobi-Bellman equation. However, the geometrical singularities of the domain and the data discontinuities do not allow us to apply the classical theory of viscosity solutions. In the first part of this thesis, we prove this kind of characterization for the value functions of optimal control problems defined on 1-dimensional networks. In the second part, the previous results are extended to the case of control problems defined on a 2-dimensional junction. Finally, in the last part, we use the results obtained previously to treat a singular perturbation problem involving optimal control problems in the plane for which the dynamics and running costs can be discontinuous through an oscillating border.

Nesta dissertação de Mestrado será apresentada uma poderosa técnica numérica, conhecida como método Level Set, capaz de simular e analisar movimentos de curvas em diferentes cenários físicos. Tal método - formulado por Osher e Sethian [1] - está sedimentado na seguinte idéia: representar uma determinada curva (ou superfície) Γ como a curva de nível zero (zero level set) de uma função Φ de maior dimensão (denominada função Level Set). A equação diferencial do tipo Hamilton-Jacobi que descreve a evolução da função Level Set é discretizada através da utilização de acurados esquemas hiperbólicos e, como resultado de tal acurácia, obtém-se uma formulação numérica capaz de tratar eficazmente mudanças topológicas e/ou descontinuidades que, eventualmente, podem surgir no decorrer da propagação da curva (ou superfície) de nível zero. Em virtude da eficácia e versatilidade do método Level Set, esta técnica numérica está sendo amplamente aplicada à diversas áreas científicas, incluindo mecânica dos fluidos, processamento de imagens e visão computacional, crescimento de cristais, geometria computacional e ciência dos materiais. Particularmente, o propósito deste trabalho equivale ao estudo dos fundamentos do método Level Set e, por fim, visa-se aplicar tal modelo numérico à problemas existentes na área de crescimento de cristais. [1] S. Osher and J. A. Sethian, Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations, J. Comp. Phys., 79:12, 1988. / In this dissertation, we present a powerful numerical technique known as Level Set Method for computing and analyzing moving fronts in different physical settings. The method -formulated by Osher and Sethian [1] - is based on the following idea: a curve (or surface) is embedded as the zero level set of a higher-dimensional function Φ (called level set function). Then, we can link the evolution of this function Φ to the propagation of the curve itself through a time-dependent initial value problem. At any time, the curve is given by the zero level set of the time-dependent level set function Φ. The evolution of the level set function Φ is described by a Hamilton-Jacobi type partial differential equation, which can be discretised by the use of accurate methods for hyperbolic equations. As a result, the Level Set Method is able to track complex curves that can develop large spikes, sharp corners or change its topology as they evolve. Because of its versatility and efficacy, this numerical technique has found applications in a large number of areas, including fluid mechanics, image processing and computer vision, crystal growth, computational geometry and materials science. Particularly, the aim of this dissertation has been to understand the fundamentals of Level Set Method and its final goal is compute the motion of bondaries in crystal growth using this numerical model. [1] S. Osher and J. A. Sethian, Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations, J. Comp. Phys., 79:12, 1988.

Nesta dissertação de Mestrado será apresentada uma poderosa técnica numérica, conhecida como método Level Set, capaz de simular e analisar movimentos de curvas em diferentes cenários físicos. Tal método - formulado por Osher e Sethian [1] - está sedimentado na seguinte idéia: representar uma determinada curva (ou superfície) Γ como a curva de nível zero (zero level set) de uma função Φ de maior dimensão (denominada função Level Set). A equação diferencial do tipo Hamilton-Jacobi que descreve a evolução da função Level Set é discretizada através da utilização de acurados esquemas hiperbólicos e, como resultado de tal acurácia, obtém-se uma formulação numérica capaz de tratar eficazmente mudanças topológicas e/ou descontinuidades que, eventualmente, podem surgir no decorrer da propagação da curva (ou superfície) de nível zero. Em virtude da eficácia e versatilidade do método Level Set, esta técnica numérica está sendo amplamente aplicada à diversas áreas científicas, incluindo mecânica dos fluidos, processamento de imagens e visão computacional, crescimento de cristais, geometria computacional e ciência dos materiais. Particularmente, o propósito deste trabalho equivale ao estudo dos fundamentos do método Level Set e, por fim, visa-se aplicar tal modelo numérico à problemas existentes na área de crescimento de cristais. [1] S. Osher and J. A. Sethian, Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations, J. Comp. Phys., 79:12, 1988. / In this dissertation, we present a powerful numerical technique known as Level Set Method for computing and analyzing moving fronts in different physical settings. The method -formulated by Osher and Sethian [1] - is based on the following idea: a curve (or surface) is embedded as the zero level set of a higher-dimensional function Φ (called level set function). Then, we can link the evolution of this function Φ to the propagation of the curve itself through a time-dependent initial value problem. At any time, the curve is given by the zero level set of the time-dependent level set function Φ. The evolution of the level set function Φ is described by a Hamilton-Jacobi type partial differential equation, which can be discretised by the use of accurate methods for hyperbolic equations. As a result, the Level Set Method is able to track complex curves that can develop large spikes, sharp corners or change its topology as they evolve. Because of its versatility and efficacy, this numerical technique has found applications in a large number of areas, including fluid mechanics, image processing and computer vision, crystal growth, computational geometry and materials science. Particularly, the aim of this dissertation has been to understand the fundamentals of Level Set Method and its final goal is compute the motion of bondaries in crystal growth using this numerical model. [1] S. Osher and J. A. Sethian, Fronts propagating with curvature-dependent speed: Algorithms based on Hamilton-Jacobi formulations, J. Comp. Phys., 79:12, 1988.

Esta dissertação aborda a teoria dos jogos diferenciais em sua estreita relação com a teoria das equações de Hamilton-Jacobi (HJ). Inicialmente, uma revisão da noção de solução em teoria dos jogos é empreendida. Discutem-se nesta ocasião as idéias de equilíbrio de Nash e alguns de seus refinamentos. Em seguida, tem lugar uma introdução à teoria dos jogos diferenciais, onde noções de solução como a função de valor de Isaacs e de Friedman são discutidas. É nesta altura do trabalho que fica evidente a conexão entre este conceito de solução e a teoria das equações de Hamilton-Jacobi. Por ocasião desta conexão, é explorada a noção de solução clássica e é exposta uma demonstração do fato de que se um jogo diferencial possuir uma função de valor pelo menos continuamente diferenciável, esta será uma solução da equação de Hamilton-Jacobi associada ao jogo. Este resultado faz uso do princípio da programação dinâmica, devido a Bellman, e cuja demonstração está presente no texto. No entanto, quando a função de valor do jogo é apenas contínua, então embora esta não seja uma solução clássica da equação HJ associada a jogo, vemos que ela será uma solução viscosa, ou solução no sentido da viscosidade - e a esta altura são discutidos os elementos e propriedades desta classe de soluções, um teorema de existência e unicidade e alguns exemplos. Por fim, retomamos o estudo dos jogos diferenciais à luz das soluções viscosas da equação de Hamilton-Jacobi e, assim, expomos uma demonstração de existência da função de valor e do princípio da programação dinâmica a partir das noções da viscosidade / This dissertation aims to address the topic of Differential Game Theory in its connection with the Hamilton-Jacobi (HJ) equations framework. Firstly we introduce the idea of solution for a game, through the discussion of Nash equilibria and its refinements. Secondly, the solution concept is then translated to the context of Differential Games and the idea of value function is introduced in its Isaacs\'s as well as Friedman\'s version. As the value function is discussed, its relationship with the Hamilton-Jacobi equations theory becomes self-evident. Due to such relation, we investigate the HJ equation from two distinct points of view. First of all, we discuss a statement according to which if a differential game has a continuously differentiable value function, then such function is a classical solution of the HJ equation associated to the game. This result strongly relies on Bellman\'s Dynamic Programming Principle - and this is the reason why we devote an entire chapter to this theme. Furthermore, HJ is still at our sight from the PDE point of view. Our motivation is simple: under some lack of regularity - a value function which is continuous, but not continuously differentiable - a game may still have a value function represented as a solution of the associated HJ equation. In this case such a solution will be called a solution in the viscosity sense. We then discuss the properties of viscosity solutions as well as provide an existence and uniqueness theorem. Finally we turn our attention back to the theory of games and - through the notion of viscosity - establish the existence and uniqueness of value functions for a differential game within viscosity solution theory.

We propose multigrid methods for solving Hamilton-Jacobi-Bellman (HJB) and Hamilton-Jacobi-Bellman-Isaacs (HJBI) equations. The methods are based on the full approximation scheme. We propose a damped-relaxation method as smoother for multigrid. In contrast with policy iteration, the relaxation scheme is convergent for both HJB and HJBI equations. We show by local Fourier analysis that the damped-relaxation smoother effectively reduces high frequency error. For problems where the control has jumps, restriction and interpolation methods are devised to capture the jump on the coarse grid as well as during coarse grid correction. We will demonstrate the effectiveness of the proposed multigrid methods for solving HJB and HJBI equations arising from option pricing as well as problems where policy iteration does not converge or converges slowly.