This thesis deals with discrete-time Markov jump linear systems (MJLS) with Markov chain in a general Borel space S. Several control issues have been addressed for this class of dynamic systems, including stochastic stability (SS), linear quadratic (LQ) optimal control synthesis, fllter design and a separation principle. Necessary and sffcient conditions for SS have been derived. It was shown that SS is equivalent to the spectral radius of an operator being less than 1 or to the existence of a solution to a \\Lyapunov-like\" equation. Based on the SS concept, the finite- and infinite-horizon LQ optimal control problems were tackled. The solution to the finite- (infinite-)horizon LQ optimal control problem was derived from the associated control S-coupled Riccati difference (algebraic) equations. By S-coupled it is meant that the equations are coupled via an integral over a transition probability kernel having a density with respect to a in-finite measure on the Borel space S. The design of linear Markov jump filters was analyzed and a solution to the finite- (infinite-)horizon filtering problem was obtained based on the associated filtering S-coupled Riccati difference (algebraic) equations. Conditions for the existence and uniqueness of a stabilizing positive semi-definite solution to the control and filtering S-coupled algebraic Riccati equations have also been derived. Finally a separation principle for discrete-time MJLS with Markov chain in a general state space was obtained. It was shown that the optimal controller for a partial information optimal control problem separates the partial information control problem into two problems, one associated with a filtering problem and the other associated with an optimal control problem with complete information. It is expected that the results obtained in this thesis may motivate further research on discrete-time MJLS with Markov chain in a general state space. / Esta tese trata de sistemas lineares com saltos markovianos (MJLS) a tempo discreto com cadeia de Markov em um espaço geral de Borel S. Vários problemas de controle foram abordados para esta classe de sistemas dinâmicos, incluindo estabilidade estocástica (SS), síntese de controle ótimo linear quadrático (LQ), projeto de filtros e um princípio da separação. Condições necessárias e suficientes para a SS foram obtidas. Foi demonstrado que SS é equivalente ao raio espectral de um operador ser menor que 1 ou à existência de uma solução para uma equação de Lyapunov. Os problemas de controle ótimo a horizonte finito e infinito foram abordados com base no conceito de SS. A solução para o problema de controle ótimo LQ a horizonte finito (infinito) foi obtida a partir das associadas equações a diferenças (algébricas) de Riccati S-acopladas de controle. Por S-acopladas entende-se que as equações são acopladas por uma integral sobre o kernel estocástico com densidade de transição em relação a uma medida in-finita no espaço de Borel S. O projeto de filtros lineares markovianos foi analisado e uma solução para o problema da filtragem a horizonte finito (infinito) foi obtida com base nas associadas equações a diferenças (algébricas) de Riccati S-acopladas de filtragem. Condições para a existência e unicidade de uma solução positiva semi-definida e estabilizável para as equações algébricas de Riccati S-acopladas associadas aos problemas de controle e filtragem também foram obtidas. Por último, foi estabelecido um princípio da separação para MJLS a tempo discreto com cadeia de Markov em um espaço de estados geral. Foi demonstrado que o controlador ótimo para um problema de controle ótimo com informação parcial separa o problema de controle com informação parcial em dois problemas, um deles associado a um problema de filtragem e o outro associado a um problema de controle ótimo com informação completa. Espera-se que os resultados obtidos nesta tese possam motivar futuras pesquisas sobre MJLS a tempo discreto com cadeia de Markov em um espaço de estados geral.

Consider a time-dependent optimal control problem, where the state evolution is described by an initial value problem. There are a variety of numerical methods to solve these problems. The so-called indirect approach is considered detailed in this thesis. The indirect methods solve decoupled boundary value problems resulting from the necessary conditions for the optimal control problem. The so-called Pantoja method describes a computationally efficient stage-wise construction of the Newton direction for the discrete-time optimal control problem. There are many relationships between multiple shooting techniques and Pantoja method, which are investigated in this thesis. In this context, the equivalence of Pantoja method and multiple shooting method of Riccati type is shown. Moreover, Pantoja method is extended to the case where the state equations are discretised using one of implicit numerical methods. Furthermore, the concept of symplecticness and Hamiltonian systems is introduced. In this regard, a suitable numerical method is presented, which can be applied to unconstrained optimal control problems. It is proved that this method is a symplectic one. The iterative solution of optimal control problems in ordinary differential equations by Pantoja or Riccati equivalent methods leads to a succession of triple sweeps through the discretised time interval. The second (adjoint) sweep relies on information from the first (original) sweep, and the third (final) sweep depends on both of them. Typically, the steps on the adjoint sweep involve more operations and require more storage than the other two. The key difficulty is given by the enormous amount of memory required for the implementation of these methods if all states throughout forward and adjoint sweeps are stored. One of goals of this thesis is to present checkpointing techniques for memory reduced implementation of these methods. For this purpose, the well known aspect of checkpointing has to be extended to a `nested checkpointing` for multiple transversals. The proposed nested reversal schedules drastically reduce the required spatial complexity. The schedules are designed to minimise the overall execution time given a certain total amount of storage for the checkpoints. The proposed scheduling schemes are applied to the memory reduced implementation of the optimal control problem of laser surface hardening and other optimal control problems. / Es wird ein Problem der optimalen Steuerung betrachtet. Die dazugehoerigen Zustandsgleichungen sind mit einer Anfangswertaufgabe definiert. Es existieren zahlreiche numerische Methoden, um Probleme der optimalen Steuerung zu loesen. Der so genannte indirekte Ansatz wird in diesen Thesen detailliert betrachtet. Die indirekten Methoden loesen das aus den Notwendigkeitsbedingungen resultierende Randwertproblem. Das so genannte Pantoja Verfahren beschreibt eine zeiteffiziente schrittweise Berechnung der Newton Richtung fuer diskrete Probleme der optimalen Steuerung. Es gibt mehrere Beziehungen zwischen den unterschiedlichen Mehrzielmethoden und dem Pantoja Verfahren, die in diesen Thesen detailliert zu untersuchen sind. In diesem Zusammenhang wird die aequivalence zwischen dem Pantoja Verfahren und der Mehrzielmethode vom Riccati Typ gezeigt. Ausserdem wird das herkoemlige Pantoja Verfahren dahingehend erweitert, dass die Zustandsgleichungen mit Hilfe einer impliziten numerischen Methode diskretisiert sind. Weiterhin wird das Symplektische Konzept eingefuehrt. In diesem Zusammenhang wird eine geeignete numerische Methode praesentiert, die fuer ein unrestringiertes Problem der optimalen Steuerung angewendet werden kann. In diesen Thesen wird bewiesen, dass diese Methode symplectisch ist. Das iterative Loesen eines Problems der optimalen Steuerung in gewoenlichen Differentialgleichungen mit Hilfe von Pantoja oder Riccati aequivalenten Verfahren fuehrt auf eine Aufeinanderfolge der Durchlaeufetripeln in einem diskretisierten Zeitintervall. Der zweite (adjungierte) Lauf haengt von der Information des ersten (primalen) Laufes, und der dritte (finale) Lauf haeng von den beiden vorherigen ab. Ueblicherweise beinhalten Schritte und Zustaende des adjungierten Laufes wesentlich mehr Operationen und benoetigen auch wesentlich mehr Speicherplatzkapazitaet als Schritte und Zustaende der anderen zwei Durchlaeufe. Das Grundproblem besteht in einer enormen Speicherplatzkapazitaet, die fuer die Implementierung dieser Methoden benutzt wird, falls alle Zustaende des primalen und des adjungierten Durchlaufes zu speichern sind. Ein Ziel dieser Thesen besteht darin, Checkpointing Strategien zu praesentieren, um diese Methoden speichereffizient zu implementieren. Diese geschachtelten Umkehrschemata sind so konstruiert, dass fuer einen gegebenen Speicherplatz die gesamte Laufzeit zur Abarbeitung des Umkehrschemas minimiert wird. Die aufgestellten Umkehrschemata wurden fuer eine speichereffiziente Implementierung von Problemen der optimalen Steuerung angewendet. Insbesondere betrifft dies das Problem einer Oberflaechenabhaertung mit Laserbehandlung.

Consider a time-dependent optimal control problem, where the state evolution is described by an initial value problem. There are a variety of numerical methods to solve these problems. The so-called indirect approach is considered detailed in this thesis. The indirect methods solve decoupled boundary value problems resulting from the necessary conditions for the optimal control problem. The so-called Pantoja method describes a computationally efficient stage-wise construction of the Newton direction for the discrete-time optimal control problem. There are many relationships between multiple shooting techniques and Pantoja method, which are investigated in this thesis. In this context, the equivalence of Pantoja method and multiple shooting method of Riccati type is shown. Moreover, Pantoja method is extended to the case where the state equations are discretised using one of implicit numerical methods. Furthermore, the concept of symplecticness and Hamiltonian systems is introduced. In this regard, a suitable numerical method is presented, which can be applied to unconstrained optimal control problems. It is proved that this method is a symplectic one. The iterative solution of optimal control problems in ordinary differential equations by Pantoja or Riccati equivalent methods leads to a succession of triple sweeps through the discretised time interval. The second (adjoint) sweep relies on information from the first (original) sweep, and the third (final) sweep depends on both of them. Typically, the steps on the adjoint sweep involve more operations and require more storage than the other two. The key difficulty is given by the enormous amount of memory required for the implementation of these methods if all states throughout forward and adjoint sweeps are stored. One of goals of this thesis is to present checkpointing techniques for memory reduced implementation of these methods. For this purpose, the well known aspect of checkpointing has to be extended to a `nested checkpointing` for multiple transversals. The proposed nested reversal schedules drastically reduce the required spatial complexity. The schedules are designed to minimise the overall execution time given a certain total amount of storage for the checkpoints. The proposed scheduling schemes are applied to the memory reduced implementation of the optimal control problem of laser surface hardening and other optimal control problems. / Es wird ein Problem der optimalen Steuerung betrachtet. Die dazugehoerigen Zustandsgleichungen sind mit einer Anfangswertaufgabe definiert. Es existieren zahlreiche numerische Methoden, um Probleme der optimalen Steuerung zu loesen. Der so genannte indirekte Ansatz wird in diesen Thesen detailliert betrachtet. Die indirekten Methoden loesen das aus den Notwendigkeitsbedingungen resultierende Randwertproblem. Das so genannte Pantoja Verfahren beschreibt eine zeiteffiziente schrittweise Berechnung der Newton Richtung fuer diskrete Probleme der optimalen Steuerung. Es gibt mehrere Beziehungen zwischen den unterschiedlichen Mehrzielmethoden und dem Pantoja Verfahren, die in diesen Thesen detailliert zu untersuchen sind. In diesem Zusammenhang wird die aequivalence zwischen dem Pantoja Verfahren und der Mehrzielmethode vom Riccati Typ gezeigt. Ausserdem wird das herkoemlige Pantoja Verfahren dahingehend erweitert, dass die Zustandsgleichungen mit Hilfe einer impliziten numerischen Methode diskretisiert sind. Weiterhin wird das Symplektische Konzept eingefuehrt. In diesem Zusammenhang wird eine geeignete numerische Methode praesentiert, die fuer ein unrestringiertes Problem der optimalen Steuerung angewendet werden kann. In diesen Thesen wird bewiesen, dass diese Methode symplectisch ist. Das iterative Loesen eines Problems der optimalen Steuerung in gewoenlichen Differentialgleichungen mit Hilfe von Pantoja oder Riccati aequivalenten Verfahren fuehrt auf eine Aufeinanderfolge der Durchlaeufetripeln in einem diskretisierten Zeitintervall. Der zweite (adjungierte) Lauf haengt von der Information des ersten (primalen) Laufes, und der dritte (finale) Lauf haeng von den beiden vorherigen ab. Ueblicherweise beinhalten Schritte und Zustaende des adjungierten Laufes wesentlich mehr Operationen und benoetigen auch wesentlich mehr Speicherplatzkapazitaet als Schritte und Zustaende der anderen zwei Durchlaeufe. Das Grundproblem besteht in einer enormen Speicherplatzkapazitaet, die fuer die Implementierung dieser Methoden benutzt wird, falls alle Zustaende des primalen und des adjungierten Durchlaufes zu speichern sind. Ein Ziel dieser Thesen besteht darin, Checkpointing Strategien zu praesentieren, um diese Methoden speichereffizient zu implementieren. Diese geschachtelten Umkehrschemata sind so konstruiert, dass fuer einen gegebenen Speicherplatz die gesamte Laufzeit zur Abarbeitung des Umkehrschemas minimiert wird. Die aufgestellten Umkehrschemata wurden fuer eine speichereffiziente Implementierung von Problemen der optimalen Steuerung angewendet. Insbesondere betrifft dies das Problem einer Oberflaechenabhaertung mit Laserbehandlung.

L'objectif de cette thèse est de développer une nouvelle méthodologie pour la détection active de défaillances, basée sur approche multimodèle et robuste des fautes. Ce travail prolonge des recherches effectuées dans le projet Metalau de l'Inria. L'apport essentiel de cette thèse est la prise en compte de modèles évoluant en boucle fermée. On utilise une approche multi-modèle pour modéliser le modèle en fonctionnement normal et le modèle défaillant. Les avantages potentiels de l'utilisation d'un feedback dynamique linéaire et ses propriétés de robustesse sont analysés dans la construction de signaux de détection auxiliaires. On compare les résultats obtenus avec ceux du cas boucle ouverte. La formulation du problème de détection active dans le cas d'un modèle en boucle fermée est nouvelle et repose sur la prise en considération de la norme du signal de détection auxiliaire comme critère d'optimisation. On considère aussi des fonctions coût plus générales, telles celles qui sont utilisées pour mesurer la performance de feedbacks dans des problèmes de la théorie de la commande linéaire robuste. La solution complète repose sur la résolution de plusieurs problèmes d'optimisation non standards / The aim is to develop a novel theory of robust active failure detection based on multi-model formulation of faults. The original method was already proposed by the Metalau group of INRIA. We have continued to work on the extension of this approach to more general cases. The focus is on the effects of feedback on the previous approach. The multi-model approach is still used to model the normal and the failed systems; however the possible advantages of using linear dynamic feedback in the construction of the auxiliary signal for robust fault detection is considered and the results are compared to the previously developed open-loop setup. An original formulation of the active fault detection problem using feedback is developed. The norm of the auxiliary signal is considered as a possible cost criterion. Also, we have considered a more general cost function that has already been used for measuring the performance of feedback configurations in Linear Control Theory. We have given a complete solution to this problem. In order to find a complete solution, several mathematical problems are solved

Esta dissertação apresenta um estudo dos filtros de Kalman para sistemas singulares em tempo discreto. Novos algoritmos são formulados para as estimativas filtradas, preditoras e suavizadas com as correspondentes equações de Riccati para sistemas singulares variantes no tempo. Nesta dissertação considera-se também uma aproximação do problema de filtragem de Kalman como um problema determinístico de ajuste ótimo de trajetória. A formulação proposta permite considerar um atraso no sinal de medida, sendo permitida a correlação entre os estados e os ruídos da medida. Apresentam-se também as provas da estabilidade e da convergência destes filtros. / This dissertation presents a study of Kalman filters for singular systems in discrete time. New algorithms are developed for the Kalman filtered, predicted and smoothed estimate recursions with the corresponding Riccati equations for time-variant singular systems. This dissertation addresses the Kalman filtering problem as a deterministic optimal trajectory fitting problem. The problem is formulated taking into account one delay in the measured signals and correlations between state and measurement noises. In the final, this work presents the stability and convergence proofs of these filters.

Esta tese de doutorado aborda os projetos robustos de controle e estimativa de estados para Sistemas Lineares sujeitos a Saltos Markovianos (SLSM) de tempo discreto sob a influência de incertezas paramétricas. Esses projetos são desenvolvidos por meio de extensões dos critérios quadráticos clássicos para SLSM nominais. Os critérios de custo quadrático para os SLSM incertos são formulados na forma de problemas de otimização min-max que permitem encontrar a melhor solução para o pior caso de incerteza (máxima influência de incerteza). Os projetos robustos correspondem às soluções ótimas obtidas por meio da combinação dos métodos de funções penalidade e mínimos quadrados regularizados robustos. Duas situações são investigadas: regular e estimar os estados quando os modos de operações são observados; e estimar os estados sob a hipótese de desconhecimento da cadeia de Markov. Estruturalmente, o regulador e as estimativas de estados assemelham-se às respectivas versões nominais. A recursividade é estabelecida em termos de equações de Riccati sem a necessidade de ajuste de parâmetros auxiliares e dependente apenas das matrizes de parâmetros e ponderações conhecidas. / This thesis deals with recursive robust designs of control and state estimates for discretetime Markovian Jump Linear Systems (MJLS) subject to parametric uncertainties. The designs are developed considering extensions of the standard quadratic cost criteria for MJLS without uncertainties. The quadratic cost criteria for uncertain MJLS are formulated in the form of min-max optimization problems to get the best solution for the worst uncertainty case. The optimal robust schemes correspond to the optimal solution obtained by the combination of penalty function and robust regularized least-squares methods. Two cases are investigated: to control and estimate the states when the operation modes are observed; and, to estimate the states when the Markov chain is unobserved. The optimal robust LQR and Kalman-type state estimates resemble the respective nominal versions. The recursiveness is established by Riccati equations in terms of parameter and weighting matrices previously known and without extra offline computations.

Esta tese aborda o problema de regulação para sistemas lineares sujeitos a saltos Markovianos de tempo discreto com matrizes de transição incertas. Considera-se que as incertezas são limitadas em norma e os estados da cadeia de Markov podem não ser completamente observados pelo controlador. No cenário com observação completa dos estados, a solução é deduzida com base em um funcional quadrático dado em termos das probabilidades de transição incertas. Enquanto que no cenário sem observação, a solução é obtida por meio da reformulação do sistema Markoviano como um sistema determinístico, independente da cadeia de Markov. Três modelos são propostos para essa reformulação: um modelo é baseado no primeiro momento do sistema Markoviano, o segundo é obtido a partir da medida de Dirac e resulta em um sistema aumentado, e o terceiro fornece um sistema aumentado singular. Os reguladores recursivos robustos são projetados a partir de critérios de custo quadrático, dados em termos de problemas de otimização restritos. A solução é derivada da técnica de mínimos quadrados regularizados robustos e apresentada em uma estrutura matricial. A recursividade é estabelecida por equações de Riccati, que se assemelham às soluções dos reguladores clássicos, para essa classe de sistemas, quando não estão sujeitos a incertezas. / This thesis deals with regulation problem for discrete-time Markovian jump linear systems with uncertain transition matrix. The uncertainties are assumed to be normbounded type. The states of the Markov chain can not be completely observed by the controller. In the scenario with complete observation of the states, the solution is deduced based on a quadratic functional given in terms of uncertain transition probabilities. While in the scenario without observation, the solution is obtained from reformulation of the Markovian system as a deterministic system, independent of the Markov chain. Three models are proposed for the reformulation process: a model is based on the first moment of the Markovian system, the second is obtained from Dirac measure which results in an augmented system, and the third provides a singular augmented system. Recursive robust regulators are designed from quadratic cost criteria given in terms of constrained optimization problems. The solution is derived from the robust regularized least-square approach, whose framework is given in terms of a matrix structure. The recursiveness is established by Riccati equations which resemble the solutions of standard regulators for this class of systems, when they are not subject to uncertainties.

Esta tese de doutorado aborda os projetos robustos de controle e estimativa de estados para Sistemas Lineares sujeitos a Saltos Markovianos (SLSM) de tempo discreto sob a influência de incertezas paramétricas. Esses projetos são desenvolvidos por meio de extensões dos critérios quadráticos clássicos para SLSM nominais. Os critérios de custo quadrático para os SLSM incertos são formulados na forma de problemas de otimização min-max que permitem encontrar a melhor solução para o pior caso de incerteza (máxima influência de incerteza). Os projetos robustos correspondem às soluções ótimas obtidas por meio da combinação dos métodos de funções penalidade e mínimos quadrados regularizados robustos. Duas situações são investigadas: regular e estimar os estados quando os modos de operações são observados; e estimar os estados sob a hipótese de desconhecimento da cadeia de Markov. Estruturalmente, o regulador e as estimativas de estados assemelham-se às respectivas versões nominais. A recursividade é estabelecida em termos de equações de Riccati sem a necessidade de ajuste de parâmetros auxiliares e dependente apenas das matrizes de parâmetros e ponderações conhecidas. / This thesis deals with recursive robust designs of control and state estimates for discretetime Markovian Jump Linear Systems (MJLS) subject to parametric uncertainties. The designs are developed considering extensions of the standard quadratic cost criteria for MJLS without uncertainties. The quadratic cost criteria for uncertain MJLS are formulated in the form of min-max optimization problems to get the best solution for the worst uncertainty case. The optimal robust schemes correspond to the optimal solution obtained by the combination of penalty function and robust regularized least-squares methods. Two cases are investigated: to control and estimate the states when the operation modes are observed; and, to estimate the states when the Markov chain is unobserved. The optimal robust LQR and Kalman-type state estimates resemble the respective nominal versions. The recursiveness is established by Riccati equations in terms of parameter and weighting matrices previously known and without extra offline computations.

Esta dissertação de mestrado aborda o problema de regulação robusta recursiva para sistemas lineares discretos sujeitos a incertezas paramétricas. Um novo funcional quadrático, baseado na combinação de função penalidade e função custo do tipo jogos, é projetado para lidar com este problema. Uma característica interessante desta abordagem é que a recursividade pode ser realizada sem a necessidade do ajuste de parâmetros auxiliares. Bastante útil para aplicações online. A solução proposta é baseada numa equação recursiva de Riccati. Também, a convergência e a estabilidade do regulador para o sistema linear incerto invariante no tempo são garantidas. / This dissertation deals with robust recursive regulators for discrete-time systems subject to parametric uncertainties. A new quadratic functional based on the combination of penalty functions and game theory is proposed to solve this class of problems. An important issue of this approach is that the recursiveness can be performed without the need of adjusting auxiliary parameters. It is useful for online applications. The solution proposed is based on Riccati equation which guarantees the convergence and stability of the time-invariant system.

Esta dissertação de mestrado aborda o problema de regulação robusta recursiva para sistemas lineares discretos sujeitos a incertezas paramétricas. Um novo funcional quadrático, baseado na combinação de função penalidade e função custo do tipo jogos, é projetado para lidar com este problema. Uma característica interessante desta abordagem é que a recursividade pode ser realizada sem a necessidade do ajuste de parâmetros auxiliares. Bastante útil para aplicações online. A solução proposta é baseada numa equação recursiva de Riccati. Também, a convergência e a estabilidade do regulador para o sistema linear incerto invariante no tempo são garantidas. / This dissertation deals with robust recursive regulators for discrete-time systems subject to parametric uncertainties. A new quadratic functional based on the combination of penalty functions and game theory is proposed to solve this class of problems. An important issue of this approach is that the recursiveness can be performed without the need of adjusting auxiliary parameters. It is useful for online applications. The solution proposed is based on Riccati equation which guarantees the convergence and stability of the time-invariant system.