Submitted by Luciana Ferreira (lucgeral@gmail.com) on 2017-02-16T11:16:03Z No. of bitstreams: 2 Dissertação - Mayk Joaquim dos Santos - 2017.pdf: 2151565 bytes, checksum: 0afafa6be7f2f9c3ee2a27ca9bf4bf24 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-02-16T11:16:36Z (GMT) No. of bitstreams: 2 Dissertação - Mayk Joaquim dos Santos - 2017.pdf: 2151565 bytes, checksum: 0afafa6be7f2f9c3ee2a27ca9bf4bf24 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-02-16T11:16:36Z (GMT). No. of bitstreams: 2 Dissertação - Mayk Joaquim dos Santos - 2017.pdf: 2151565 bytes, checksum: 0afafa6be7f2f9c3ee2a27ca9bf4bf24 (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-01-16 / In this work we will show that, given a piecewise smooth vector field, we can apply the regularization method and, from it, via blow-up, turn it into a singular perturbation problem. By doing that, we can use the tools from singular perturbation theory to perform a qualitative study of piecewise smooth vector fields. Finally, we will show that, through successive changes of coordinates, a singularity of a discontinuous submanifold of codimension k, where k=1 or k=2, can be transformed into a singularity of codimension 0 in order to study the qualitative behavior in this submanifold, where the Filippov’s convention holds. / Neste trabalho mostraremos que, dado um campo de vetores suaves por partes, podemos aplicar o método de regularização e, a partir deste, via “blow-up”, o transformamos em um problema de perturbação singular. Podemos, dessa forma, fazer uso das ferramentas da teoria de perturbação singular para realizar um estudo qualitativo dos campos de vetores suaves por partes. Por último, mostraremos que através de sucessivas mudanças de coordenadas podemos transformar uma singularidade de uma subvariedade de descontinuidade de codimensão k, onde k=1 ou k=2, em uma uma singularidade de codimensão 0 e estudar o comportamento qualitativo ao longo desta subvariedade, onde é válida a convenção de Filippov.

Submitted by Luciana Ferreira (lucgeral@gmail.com) on 2017-03-22T12:19:52Z No. of bitstreams: 2 Tese - Oscar Alexander Ramírez Cespedes - 2017.pdf: 14139665 bytes, checksum: 57b8ba9047422f62ed013be3b7bf660e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-03-22T12:20:24Z (GMT) No. of bitstreams: 2 Tese - Oscar Alexander Ramírez Cespedes - 2017.pdf: 14139665 bytes, checksum: 57b8ba9047422f62ed013be3b7bf660e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-03-22T12:20:24Z (GMT). No. of bitstreams: 2 Tese - Oscar Alexander Ramírez Cespedes - 2017.pdf: 14139665 bytes, checksum: 57b8ba9047422f62ed013be3b7bf660e (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-03-07 / Fundação de Amparo à Pesquisa do Estado de Goiás - FAPEG / In this work, we analize the version of Hilbert’s 16th problem for a piecewise linear differential system, PWLS, in R2. More precisely,we determinete the maximum number of certain types of limit cycles when the system is define in two zones separated by a straight line. Some results on the maximum number of cycles of a PWLS defined in two sectors were established. In addition, we classify typical singularities of a piecewise smooth systemin R3, taking into account the behavior of the associated sliding field. / Neste trabalho, estudamos uma versão do 16◦ Problema de Hilbert para sistemas de equações diferenciais lineares por partes, PWLS, em R2. Mais precisamente, determinamos o número máximo de certos tipos de ciclos limite quando o sistema está definido em duas zonas separadas por uma linha reta. Alguns resultados sobreo número de máximo de ciclos de um PWLS definido em dois setores são estabelecidos. Além disso, estudamos e classificamos singularidades típicas de um sistema suave por partes em R3 levando em consideração o comportamento do campo deslizante associado.

Submitted by JÚLIO HEBER SILVA (julioheber@yahoo.com.br) on 2017-06-13T19:33:36Z No. of bitstreams: 2 Tese - Bruno Rodrigues de Freitas - 2016.pdf: 2506265 bytes, checksum: be3374b7af568ce914be02e5fa39c4ad (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-06-14T15:23:37Z (GMT) No. of bitstreams: 2 Tese - Bruno Rodrigues de Freitas - 2016.pdf: 2506265 bytes, checksum: be3374b7af568ce914be02e5fa39c4ad (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-06-14T15:23:38Z (GMT). No. of bitstreams: 2 Tese - Bruno Rodrigues de Freitas - 2016.pdf: 2506265 bytes, checksum: be3374b7af568ce914be02e5fa39c4ad (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2016-05-13 / Fundação de Amparo à Pesquisa do Estado de Goiás - FAPEG / We consider a class of piecewise linear di erential systems in R3 separated by a plane and we study its global and local dynamics. More precisely, we give conditions to the existence of invariant surfaces and limit cycles, presenting the maximum number of limit cycles and characterizing these invariant surfaces. Also, we obtain results about the T-singularity obtained by a perturbation of piecewise linear di erential systems. In our approach, we use many techniques, as an extension of the theorem’s Rolle for vector fields, Theory of Sturm’s sequence, extendedcomplete Tchebyche systems and extensions of Averaging theory. / Consideramos uma classe de sistemas diferenciais lineares por partes em R3 separados por umplano e estudamos sua dinâmica global e local. Mais precisamente, damos condições para a existência de superfícies invariantes e ciclos limite, apresentando o número máximo de ciclos limite e caracterizando estas superfícies. Obtemos resultados sobre a T-singularidade obtida por uma perturbação de sistemas diferenciais lineares por partes. Em nossa abordagem, usamos várias técnicas como uma extensão do teorema de Rolle para campos, teoria da sequência de Sturm, sistemas estendidos completos de Tchebyche e extensões da teoria da Média.

Nessa tese, estudamos o limite hidrodinâmico de um sistema estocástico de neurônios cujas interações são dadas por potenciais de Kac que imitam sinapses elétricas e químicas, e as correntes de vazamento. Esse sistema consiste de $\\ep^$ neurônios imersos em $[0,1)^2$, cada um disparando aleatoriamente de acordo com um processo pontual com taxa que depende tanto do seu potential de membrana como da posição. Quando o neurônio $i$ dispara, seu potential de membrana é resetado para $0$, enquanto que o potencial de membrana do neurônio $j$ é aumentado por um valor positivo $\\ep^2 a(i,j)$, se $i$ influencia $j$. Além disso, entre disparos consecutivos, o sistema segue uma movimento determinístico devido às sinapses elétricas e às correntes de vazamento. As sinapses elétricas estão envolvidas na sincronização do potencial de membrana dos neurônios, enquanto que as correntes de vazamento inibem a atividade de todos os neurônios, atraindo simultaneamente todos os potenciais de membrana para $0$. No principal resultado dessa tese, mostramos que a distribuição empírica dos potenciais de membrana converge, quando o parâmetro $\\ep$ tende à 0 , para uma densidade de probabilidade $ho_t(u,r)$ que satisfaz uma equação diferencial parcial nâo linear do tipo hiperbólica . / We study the hydrodynamic limit of a stochastic system of neurons whose interactions are given by Kac Potentials that mimic chemical and electrical synapses and leak currents. The system consists of $\\ep^$ neurons embedded in $[0,1)^2$, each spiking randomly according to a point process with rate depending on both its membrane potential and position. When neuron $i$ spikes, its membrane potential is reset to $0$ while the membrane potential of $j$ is increased by a positive value $\\ep^2 a(i,j)$, if $i$ influences $j$. Furthermore, between consecutive spikes, the system follows a deterministic motion due both to electrical synapses and leak currents. The electrical synapses are involved in the synchronization of the membrane potentials of the neurons, while the leak currents inhibit the activity of all neurons, attracting simultaneously their membrane potentials to 0. We show that the empirical distribution of the membrane potentials converges, as $\\ep$ vanishes, to a probability density $ho_t(u,r)$ which is proved to obey a nonlinear PDE of Hyperbolic type.

Apresentamos um estudo das álgebras de incidência que são hereditárias por partes, as quais denominamos Phias, piecewise hereditary incidence algebras. Através da aljava com relações, descrevemos as Phias de tipo Dynkin e introduzimos uma nova família de Phias de tipo Dynkin extendido chamada família ANS, em referência a Assem, Nehring e Skowronski. Nessa descrição, o importante método foi o dos cortes em extensões triviais, os quais inspiraram a elaboração de um programa que concebe exatamente os cortes na extensão trivial dada que resultam em álgebras de incidência. Abordamos as Phias &#922\\&#916 de tipo feixes, estudando o &#922\\&#916-módulo sincero canônico M e a álgebra de extensão por um ponto &#922\\&#916[&#924]. Demonstramos que se &#922Q/I é uma álgebra sincera, quase-inclinada canônica de tipo aljava e tipo de representação infinito, então os &#922Q/I-módulos sinceros são excepcionais. Essa conclusão permite construir uma gama de Phias &#922\\&#916[&#924] de tipo selvagem. Exploramos as Phias simplesmente conexas, provando uma resposta positiva para o problema de Skowronski para &#922\\&#916 uma Phia de tipo H, com grafo de objetos inclinantes &#922_D^b (&#919) conexo: o grupo &#919^1(&#922\\&#916) é trivial se, e somente se, a álgebra &#922\\&#916 é simplesmente conexa. Na área homológica, determinamos um limitante superior da dimensão global forte das Phias; mais ainda, ampliamos esse resultado para as álgebras sinceras provando que dada uma álgebra sincera e hereditária por partes, sua dimensão global forte é menor ou igual a três. / We present a study of incidence algebras that are piecewise hereditary, which we denominate Phias. By means of the quiver with relations, we describe Phias of Dynkin type and introduce a new family of Phias of extended Dynkin type, which we call ANS family, in reference to Assem, Nehring, and Skowronski. In this description, the important method was the one of cuts on trivial extensions, inspiring the writing of a program that shows exactly the cuts on the given trivial extension that result on incidence algebras. We approach sheaves type Phias &#922\\&#916, studying the canonical sincere &#922\\&#916-module M and the one-point extension algebra &#922\\&#916[&#924]. We show that if &#922Q/I is a sincere, quasi-tilted canonical algebra of quiver type and infinite representation type, then sincere &#922Q/I-modules are exceptional. This conclusion allows the construction of a wide range of Phias &#922\\&#916[&#924] wild type. We explore the simply conectedeness of Phias, proving a positive answer of the so called Skowronski problem for &#922\\&#916 a Phia H type, with connected quiver of tilting objects &#922_D^b (&#919): the group &#919^1(&#922\\&#916) is trivial if, and only if, &#922\\&#916 is a simply connected algebra. On homology, we determine an upper bound for the strong global dimension of Phias; furthermore, we extend this result for sincere algebras proving that the strong global dimension of a sincere piecewise hereditary algebra is less or equal to three.

Ce travail de thèse est consacré à l’étude d’équations d’évolution et d’équations différentielles à argument constant par morceaux. L’étude des équations différentielles à argument constant par morceaux est un domaine important car ces équations ont la structure de système dynanmique de longueur constante. La continuité des solutions conduit à une relation de récurrence entre les valeurs de cette dernière entre les points n et n+1, où n est un entier relatif quelconque. Par conséquent les équations différentielles à argument constant par morceaux combinent à la fois les propriétés des équations différentielles et des équations aux différences. Nous étudierons l’existence de solutions presque automorphes et S-asymptotiquement omega-périodiques d’équations d’évolutions et d’équations à argument constant par morceaux. L’étude de solutions presque automorphes et S’asymptotiquement omega periodiques est motivé par le fait que ces fonctions généralisent celle des fonctions périodiques. Nous obtiendrons donc des résultats concernant l’existence et l’unicité de solutions presque automorphes et S asymptotiquement omega périodiques de plusieurs équations d’évolutions. Cette problématique sera notamment étudiée dans le cadre des équations d’évolutions appartenant à la classe des équations différentielles à argument constant par morceaux. / This thesis deals with the study of evolution equations and differential equations with piecewise constant argument. Studies of such equations were motivated by the fact that they represent a hybrid of discrete and continuous dynamical systems and combine the properties of both differential and differential-difference equations. We study the existence of almost automorphic solutions and S asymptotically omega periodic solution of evolution equations and differential equations with piecewise constant argument. The study of almost automorphic and S asymptotically omega periodic functions is motivated by the fact that these functions generalize the concept of periodic functions. Therefore, we obtain results about existence and unicity of almost automorphic and S asymptotic omega periodic solution of evolution equations. We will study this problem considering evolution equations who belong to a class of differential equation with piecewise constant argument.

Ces travaux portent sur la commande et l’analyse de la stabilité d’un système électropneumatique constitué d’un axe linéaire commandé par deux servodistributeurs régulant le débit massique entrant dans chaque chambre de l’actionneur. La problématique générale est motivée par l’apparition d’un phénomène de redécollage sur ce système électropneumatique difficilement pris en compte par les études actuelles en automatique. Ce problème, rencontré depuis de nombreuses années, concerne toutes les commandes linéaires et non linéaires mono et multidimensionnelles étudiées au laboratoire. Il se traduit par des mouvements saccadés du vérin au voisinage de l’équilibre. Ce phénomène est dû à la présence de frottements secs et aux dynamiques des pressions dans les chambres pneumatiques de l’actionneur, qui continuent à évoluer (intégrer le débit massique entrant délivré par les servodistributeurs), même après l’équilibre mécanique. La première partie de ce mémoire propose une commande non linéaire commutée afin d’éviter le phénomène de redécollage de l’actionneur électropneumatique notamment vis-à-vis des variations de frottements secs qui peuvent à tout moment causer ce phénomène. Cette technique est finalement mise en œuvre et son efficacité est constatée. La plus grande partie de ce mémoire traite l’analyse de l’actionneur électropneumatique avec sa loi de commande commutée. La présence de frottements secs et l’application d’une loi de commande commutée nous a amené à concilier une démarche d’analyse de stabilité, en considérant une classe de systèmes commutés appelée systèmes affines par morceaux. La principale difficulté de cette démarche réside dans l’obtention de fonctions de Lyapunov adéquates, qui se transforme en un problème d’optimisation sous contraintes LMI (Linear Matrix Inequality) en utilisant la S-procédure. Afin d’analyser la stabilité d’un système PWA (PieceWise Affine), la première démarche proposée permet le calcul d’une fonction de Lyapunov quadratique par morceaux sous la forme d’un problème d’optimisation sous contraintes LMI, en imposant des conditions suffisantes de stabilité. Ces dernières permettent, contrairement aux méthodes classiques, d’assurer la convergence de trajectoires d’état non pas vers un point d’équilibre, mais vers un ensemble des points d’équilibre d’un système PWA. L’approche proposée permet aussi l’étude de la robustesse vis-à-vis des variations paramétriques dans le système. Nous proposons aussi une deuxième approche pour la construction d’un type de fonctions de Lyapunov dites polynomiales par morceaux, via l’utilisation des "sum of square" et de la "power transformation", afin d’analyser la stabilité d’un ensemble de points d’équilibre d’un système PWA, en présence de phénomènes de glissement et de variations paramétriques. Cette approche propose des conditions suffisantes moins conservatives que celles imposées par les fonctions de Lyapunov quadratique par morceaux. En effet, sur des exemples de systèmes PWA présentant de dynamiques discontinues sur les frontières entre les cellules, pouvant générer à tout moment des phénomènes de glissement, ces dernières s’avèrent inefficaces et ne permettent pas d’assurer la stabilité des systèmes PWA en présence de ces phénomènes. Par conséquent, les résultats sur la fonction de Lyapunov quadratique par morceaux sont étendus pour pouvoir calculer des fonctions de Lyapunov polynomiales par morceaux d’ordre supérieur, en résolvant un problème d’optimisation sous contraintes LMI. Ces dernières permettent de garantir des conditions plus générales et moins conservatives par rapport à celles développées dans la littérature. Ces deux approches ont été appliquées afin d’analyser la stabilité de l’ensemble des points d’équilibre du système électropneumatique, en considérant à la fois un modèle de frottements sous la forme d’une saturation et un autre sous la forme d’un relais présentant une dynamique discontinue. [...] / This work focuses on the control and stability analysis of an electro-pneumatic system, i.e. a linear pneumatic cylinder controlled by two servo valves regulating the mass flow entering each chamber of the actuator. The general problem is motivated by the appearance of stick-slip on the electro-pneumatic system, hardly taken into account by the current studies in automatic control. This problem, encountered throughout the years, concerns all mono- and multidimensional linear and non-linear controls systems studied at the laboratory. In pneumatic cylinders, the phenomenon consists in a displacement of the rod a while after it has come to a rest ; this is due to the fact that the force acting on the rod initially becomes smaller that the threshold which is necessary for a motion, and then this threshold is overcome later on. In this case, stick-slip is caused by the presence of dry friction and by the pressure dynamics in the chambers, which continue to evolve (integrating the net incoming mass flow from the servovalves) even after the rod has stopped. The first part of this thesis proposes a nonlinear switching control law in order to avoid stick-slip on pneumatic cylinder, taking into account with the variations of dry friction that may occur at any time causing this phenomenon. This technique is implemented and its effectiveness is recognized. The greatest part of this thesis deals with the stability analysis of the pneumatic cylinder with its switched control law. The presence of dry friction and the application of a switched control law requires an appropriate method for approaching the stability analysis ; this method is based on considering the closed-loop system as belonging to a class of switched systems called piecewise affine systems (PWA). The main difficulty in this approach lies in obtaining adequate Lyapunov functions for proving stability, which turns into an optimization problem under LMI constraints (Linear Matrix Inequality) using the S-procedure. In order to analyze the stability of a PWA system, a first method is proposed allowing the computation of a piecewise quadratic Lyapunov function through an optimization problem under LMI constraints. The methods takes into account, in contrast to conventional methods, that the states might converge not to a single point but to a set of equilibrium points. The proposed approach allows also the study of robustness with respect to parametric variations in the system. A second method is also proposed for the construction of a type of Lyapunov functions called piecewise polynomial, using the “sum of squares” and “power transformation” techniques. This approach proposes less conservative sufficient conditions than those imposed by the piecewise quadratic Lyapunov functions, yielding a more succesfull stability test when for PWA systems featuring sliding modes and parametric variations. In fact, on PWA systems with discontinuous dynamics (which can generate sliding phenomena), piecewise quadratic Lyapunov functions might prove ineffective to prove the stability. Therefore, the results on piecewise quadratic Lyapunov functions are extended in order to compute piecewise polynomial Lyapunov functions of higher order, by solving an optimization problem under LMI constraints. These functions are more general and allow less conservative conditions compared to those formerly developed in the literature. Both of these methods have been applied to the stability analysis of the set of equilibrium points of the pneumatic cylinder, considering first a friction model in saturation form and then a model in relay form with a discontinuous dynamics. The application of the methods is successful, i.e. the robust stability is proven under dry friction threshold variations, with possibility of sliding modes.

Les travaux de cette thèse portent sur l'identification des systèmes et l'extraction de motifs à partir de données. Dans le cadre de l'identification des systèmes, nous nous intéressons plus précisément à l'identification des systèmes dynamiques hybrides et l'identification des systèmes dynamiques linéaires ayant une sortie binaire. Deux classes très populaires des systèmes hybrides sont les systèmes linéaires à commutations et les systèmes affines par morceaux. Nous faisons tout d'abord un état de l'art sur les méthodes d'identification de ces deux classes. Nous proposons ensuite un algorithme basé sur une méthode d'identification de type OBE "Outer Bounding Ellipsoid" pour l'identification en temps réel des systèmes à commutations soumis à un bruit borné. Nous présentons ensuite plusieurs extensions de l'algorithme soit pour l'identification des systèmes affines par morceaux, l'identification des systèmes à commutations décrits par un modèle du type erreur de sortie et l'identification des systèmes MIMO à commutations. Nous abordons ensuite le problème d'identification des systèmes linéaires ayant une sortie binaire en introduisant un point de vue original consiste à formuler le problème d'identification comme un problème de classification. Ceci permet de proposer deux algorithmes d'identification basés sur l'utilisation des SVMs. Le premier algorithme est dédié à l'identification des systèmes à temps discret et le deuxième algorithme est dédié à l'identification des systèmes à temps continu. Dans le cadre de l'extraction de motifs, nous présentons dans un premier temps un état de l'art sur les algorithmes d'extraction de motifs et sur les techniques de la classification non supervisée. Ensuite, nous proposons un algorithme d'extraction de motifs à partir des données basé sur des techniques de classification non supervisée. / In this thesis, we deal with the identification of systems and the extraction of patterns from data. In the context of system identification, we focus precisely on the identification of hybrid systems and the identification of linear systems using binary sensors. Two very popular classes of hybrid systems are switched linear systems and piecewise affine systems. First, we give an overview of the different approaches available in the literature for the identification of these two classes. Then, we propose a new real-time identification algorithm for switched linear systems, it's based on an Outer Bounding Ellipsoid (OBE) type algorithm suitable for system identification with bounded noise. We then present several extensions of the algorithm either for the identification of piecewise affine systems, the identification of switched linear systems described by an output error model and the identification of MIMO switched linear systems. After this, we address the problem of the identification of linear systems using binary sensors by introducing an original point of view. We formulate the identification problem as a classification problem. This formulation allows the use of supervised learning algorithms such as Support Vector Machines (SVMs) for the identification of discrete time systems and the identification of continuous-time systems using binary sensors. In the context of pattern extraction, we first present an overview of the different pattern extraction algorithms and clustering techniques available in the literature. Next, we propose an algorithm for extracting patterns from data based on clustering techniques.

[pt] O objetivo desta dissertação é propor um mecanismo de testes para a avaliação dos resultados obtidos em uma modelagem TS-TARX.A principal motivação é encontrar uma solução para um problema comum na modelagem TS-TARX : os modelos espúrios que são gerados durante o processo de divisão do espaço das variáveis independentes.O modelo é uma heurística baseada em análise de árvore de regressão, como discutido por Brieman -3, 1984-. O modelo proposto para a análise de séries temporais é chamado TARX - Threshold Autoregressive with eXternal variables-. A idéia central é encontrar limiares que separem regimes que podem ser explicados através de modelos lineares. Este processo é um algoritmo que preserva o método de regressão por mínimos quadrados recursivo -MQR-. Combinando a árvore de decisão com a técnica de regressão -MQR-, o modelo se tornou o TS-TARX -Tree Structured - Threshold AutoRegression with external variables-.Será estendido aqui o trabalho iniciado por Aranha em -1, 2001-. Onde a partir de uma base de dados conhecida, um algoritmo eficiente gera uma árvore de decisão por meio de regras, e as equações de regressão estimadas para cada um dos regimes encontrados. Este procedimento pode gerar alguns modelos espúrios ou por construção,devido a divisão binária da árvore, ou pelo fato de não existir neste momento uma metodologia de comparação dos modelos resultantes.Será proposta uma metodologia através de sucessivos testes de Chow -5, 1960- que identificará modelos espúrios e reduzirá a quantidade de regimes encontrados, e consequentemente de parâmetros a estimar. A complexidade do modelo final gerado é reduzida a partir da identificação de redundâncias, sem perder o poder preditivo dos modelos TS-TARX .O trabalho conclui com exemplos ilustrativos e algumas aplicações em bases de dados sintéticas, e casos reais que auxiliarão o entendimento. / [en] The goal of this dissertation is to propose a test mechanism to evaluate the results obtained from the TS-TARX modeling procedure.The main motivation is to find a solution to a usual problem related to TS-TARX modeling: spurious models are generated in the process of dividing the space state of the independent variables.The model is a heuristics based on regression tree analysis, as discussed by Brieman -3, 1984-. The model used to estimate the parameters of the time series is a TARX -Threshold Autoregressive with eXternal variables-.The main idea is to find thresholds that split the independent variable space into regimes which can be described by a local linear model. In this process, the recursive least square regression model is preserved. From the combination of regression tree analysis and recursive least square regression techniques, the model becomes TS-TARX -Tree Structured - Threshold Autoregression with eXternal variables-.The works initiated by Aranha in -1, 2001- will be extended. In his works, from a given data base, one efficient algorithm generates a decision tree based on splitting rules, and the corresponding regression equations for each one of the regimes found.Spurious models may be generated either from its building procedure, or from the fact that a procedure to compare the resulting models had not been proposed.To fill this gap, a methodology will be proposed. In accordance with the statistical tests proposed by Chow in -5, 196-, a series of consecutive tests will be performed.The Chow tests will provide the tools to identify spurious models and to reduce the number of regimes found. The complexity of the final model, and the number of parameters to estimate are therefore reduced by the identification and elimination of redundancies, without bringing risks to the TS-TARX model predictive power.This work is concluded with illustrative examples and some applications to real data that will help the readers understanding.

On s’intéresse au problème de contrôle impulsionnel à horizon infini avec facteur d’oubli pour les processus de Markov déterministes par morceaux (PDMP). Dans un premier temps, on modélise l’évolution d’un système opto-électronique par des PDMP. Afin d’optimiser la maintenance du système, on met en place un problème de contrôle impulsionnel tenant compte à la fois du coût de maintenance et du coût lié à l’indisponibilité du matériel auprès du client.On applique ensuite une méthode d’approximation numérique de la fonction valeur associée au problème, faisant intervenir la quantification de PDMP. On discute alors de l’influence des paramètres sur le résultat obtenu. Dans un second temps, on prolonge l’étude théorique du problème de contrôle impulsionnel en construisant de manière explicite une famille de stratégies є-optimales. Cette construction se base sur l’itération d’un opérateur dit de simple-saut-ou-intervention associé au PDMP, dont l’idée repose sur le procédé utilisé par U.S. Gugerli pour la construction de temps d’arrêt є-optimaux. Néanmoins, déterminer la meilleure position après chaque intervention complique significativement la construction de telles stratégies et nécessite l’introduction d’un nouvel opérateur. L’originalité de la construction de stratégies є-optimales présentée ici est d’être explicite, au sens où elle ne nécessite pas la résolution préalable de problèmes complexes. / We are interested in a discounted impulse control problem with infinite horizon forpiecewise deterministic Markov processes (PDMPs). In the first part, we model the evolutionof an optronic system by PDMPs. To optimize the maintenance of this equipment, we study animpulse control problem where both maintenance costs and the unavailability cost for the clientare considered. We next apply a numerical method for the approximation of the value function associated with the impulse control problem, which relies on quantization of PDMPs. The influence of the parameters on the numerical results is discussed. In the second part, we extendthe theoretical study of the impulse control problem by explicitly building a family of є-optimalstrategies. This approach is based on the iteration of a single-jump-or-intervention operator associatedto the PDMP and relies on the theory for optimal stopping of a piecewise-deterministic Markov process by U.S. Gugerli. In the present situation, the main difficulty consists in approximating the best position after the interventions, which is done by introducing a new operator.The originality of the proposed approach is the construction of є-optimal strategies that areexplicit, since they do not require preliminary resolutions of complex problems.