Este trabalho discute o projeto de compensadores multivariáveis robustos com especificações no domínio do tempo. Primeiramente faz-se a análise dos compensadores por observadores de estados como forma de atingir tais objetivos. Mostra-se que, em certas condições, essa estrutura equivale à dos observadores proporcionais-integrais e apresentam-se as condições de estabilidade nominal. Evidencia-se também que é possível tratar esse problema de controle como um problema de "model matching" ou como um problema de controle com dois graus de liberdade. Mostra-se também que o projeto do compensador é equivalente ao projeto de sistemas de controle por realimentação estática da saída. Essa equivalência implica que, embora os compensadores por observadores sejam cômodos à incorporação de especificações temporais, sua estrutura é limitada para garantir que especificações gerais sejam satisfeitas. Contorna-se então essa limitação estendendo-se o estudo ao caso dos compensadores sem essa restrição estrutural. O problema de "model matching" e o problema de controle 2D são considerados como forma indireta de incorporar-se as especificações temporais e condições de projeto são obtidas reduzindo-se os possíveis conservadorismos dos projetos usuais. Ainda neste sentido, formula-se o problema denominado de "model tracking", podendo ser capaz de reduzir ainda mais esses aspectos de conservadorismo. Nessa estrutura, o erro de rastreamento entre a saída do modelo de referência e a saída medida da planta é realimentado, fazendo com que o modelo de referência faça parte explícita do controlador, permitindo ajuste pós-projeto da mesma forma que faz com a estrutura 2D. Toda as condições de projeto são obtidas no domínio da freqüência, em termos de restrições da malha aberta e da malha fechada, permitindo que técnicas padrões de projeto multivariável como LQG/LTR e H¥ possam ser diretamente utilizadas na obtenção do compensador. Obtiveram-se assim metodologias de projeto capazes de garantir robustamente a rejeição do erro de rastreamento de um modelo de referência, além de permitir que especificações usuais como a rejeição dos erros de medida e rejeição dos distúrbios sejam incorporadas no mesmo procedimento de projeto. O controle de um tanque de mistura é usado como exemplo numérico para ilustrar as metodologias de projeto. / This work is concerned with the design of robust multivariable controllers with time domain specifications. As a first step an analysis of controllers based on state observers has been done. It has been shown that this structure is equivalent to the proportional-integral observers. The conditions for nominal stability have then been presented. It has also been shown that this control problem can be seen as a model matching control design with two degrees of freedom. Furthermore it has been shown that compensator design can be reduced to the traditional static output feedback problem. This fact implies that although compensators based on state observers seem to be suitable to deal with time domain specifications, their structure has limitations to guarantee that more general specifications be satisfied. In order to overcome such limitations both the model matching and the 2-D control structures have been considered to include time domain specifications in the design problem. Design conditions have been derived in order to reduce the eventual conservatism associated to the usual design procedure. A control design problem named model tracking has also been formulated with the aim of reducing even more such conservatism. In this new structure the tracking error between the output of the reference model and the measured plant output is fed back. The reference model is then an explicit part of the controller. As a consequence small adjustments can be done on it during system start-up in the same way as with the 2-D structure. All the design conditions have been written in the frequency domain as constraints on both open-loop and closed-loop transfer matrices. Standard multivariable design techniques like LQG/LTR and H¥ can then be used. The proposed design methodology can robustly guarantee tracking of the reference model output as well as both measurement error and disturbance rejections. The control of a mixture tank has been used as a numeric example to illustrate the design methodologies.

Este trabalho discute o projeto de compensadores multivariáveis robustos com especificações no domínio do tempo. Primeiramente faz-se a análise dos compensadores por observadores de estados como forma de atingir tais objetivos. Mostra-se que, em certas condições, essa estrutura equivale à dos observadores proporcionais-integrais e apresentam-se as condições de estabilidade nominal. Evidencia-se também que é possível tratar esse problema de controle como um problema de "model matching" ou como um problema de controle com dois graus de liberdade. Mostra-se também que o projeto do compensador é equivalente ao projeto de sistemas de controle por realimentação estática da saída. Essa equivalência implica que, embora os compensadores por observadores sejam cômodos à incorporação de especificações temporais, sua estrutura é limitada para garantir que especificações gerais sejam satisfeitas. Contorna-se então essa limitação estendendo-se o estudo ao caso dos compensadores sem essa restrição estrutural. O problema de "model matching" e o problema de controle 2D são considerados como forma indireta de incorporar-se as especificações temporais e condições de projeto são obtidas reduzindo-se os possíveis conservadorismos dos projetos usuais. Ainda neste sentido, formula-se o problema denominado de "model tracking", podendo ser capaz de reduzir ainda mais esses aspectos de conservadorismo. Nessa estrutura, o erro de rastreamento entre a saída do modelo de referência e a saída medida da planta é realimentado, fazendo com que o modelo de referência faça parte explícita do controlador, permitindo ajuste pós-projeto da mesma forma que faz com a estrutura 2D. Toda as condições de projeto são obtidas no domínio da freqüência, em termos de restrições da malha aberta e da malha fechada, permitindo que técnicas padrões de projeto multivariável como LQG/LTR e H¥ possam ser diretamente utilizadas na obtenção do compensador. Obtiveram-se assim metodologias de projeto capazes de garantir robustamente a rejeição do erro de rastreamento de um modelo de referência, além de permitir que especificações usuais como a rejeição dos erros de medida e rejeição dos distúrbios sejam incorporadas no mesmo procedimento de projeto. O controle de um tanque de mistura é usado como exemplo numérico para ilustrar as metodologias de projeto. / This work is concerned with the design of robust multivariable controllers with time domain specifications. As a first step an analysis of controllers based on state observers has been done. It has been shown that this structure is equivalent to the proportional-integral observers. The conditions for nominal stability have then been presented. It has also been shown that this control problem can be seen as a model matching control design with two degrees of freedom. Furthermore it has been shown that compensator design can be reduced to the traditional static output feedback problem. This fact implies that although compensators based on state observers seem to be suitable to deal with time domain specifications, their structure has limitations to guarantee that more general specifications be satisfied. In order to overcome such limitations both the model matching and the 2-D control structures have been considered to include time domain specifications in the design problem. Design conditions have been derived in order to reduce the eventual conservatism associated to the usual design procedure. A control design problem named model tracking has also been formulated with the aim of reducing even more such conservatism. In this new structure the tracking error between the output of the reference model and the measured plant output is fed back. The reference model is then an explicit part of the controller. As a consequence small adjustments can be done on it during system start-up in the same way as with the 2-D structure. All the design conditions have been written in the frequency domain as constraints on both open-loop and closed-loop transfer matrices. Standard multivariable design techniques like LQG/LTR and H¥ can then be used. The proposed design methodology can robustly guarantee tracking of the reference model output as well as both measurement error and disturbance rejections. The control of a mixture tank has been used as a numeric example to illustrate the design methodologies.

This dissertation contains two parts: Commutative and Noncommutative Multidimensional ($d$-D) Linear Systems Theory. The first part focuses on the development of the interpolation theory to solve the $H^{\infty}$ control problem for $d$-D linear systems. We first review the classical discrete-time 1D linear system in the operator theoretical viewpoint followed by the formulations of the so-called Givone-Roesser and Fornasini-Marchesini models. Application of the $d$-variable $Z$-transform to the system of equations yields the transfer function which is a rational function of several complex variables, say $\mathbf{z} = (z_{1}, \dots, z_{d})$. We then consider the output feedback stabilization problem for a plant $P(\mathbf{z})$. By assuming that $P(\mathbf{z})$ admits a double coprime factorization, then a set of stabilizing controllers $K(\mathbf{z})$ can be parametrized by the Youla parameter $Q(\mathbf{z})$. By doing so, one can convert such a problem to the model matching problem with performance index $F(\mathbf{z})$, affine in $Q(\mathbf{z})$. Then, with $F(\mathbf{z})$ as the design parameter rather than $Q(\mathbf{z})$, one has an interpolation problem for $F(\mathbf{z})$. Incorporation of a tolerance level on $F(\mathbf{z})$ then leads to an interpolation problem of multivariable Nevanlinna-Pick type. We also give an operator-theoretic formulation of the model matching problem which lends itself to a solution via the commutant lifting theorem on the polydisk. The second part details a system whose time-axis is described by a free semigroup $\mathcal{F}_{d}$. Such a system can be represented by the so-called noncommutative Givone-Roesser, or noncommutative Fornasini-Marchesini models which are analogous to those in the first part. Application of a noncommutative $d$-variable $Z$-transform to the system of equations yields the transfer function expressed by a formal power series in several noncommuting indeterminants, say $T(z) = \sum_{v \in \mathcal{F}_{d}}T_{v}z^{v}$ where $z^{v} = z_{i_{n}} \dotsm z_{i_{1}}$ if $v = g_{i_{n}} \dotsm g_{i_{1}} \in \mathcal{F}_{d}$ and $z_{i}z_{j} \neq z_{j}z_{i}$ unless $i = j$. The concepts of reachability, controllability, observability, similarity, and stability are introduced by means of the state-space interpretation. Minimal realization problems for noncommutative Givone-Roesser or Fornasini-Marchesini systems are solved directly by a shift-realization procedure constructed from appropriate noncommutative Hankel matrices. This procedure adapts the ideas of Schützenberger and Fliess originally developed for "recognizable series" to our systems. / Ph. D.

L'Ingénierie Dirigée par les Modèles (IDM) est un domaine de recherche en pleine émergence qui considère les modèles comme des éléments de base. Chaque modèle est conforme à un autre modèle, appelé son méta-modèle, qui définit sa syntaxe abstraite et ses concepts. Dans un processus IDM, différents types de modèles sont manipulés par des transformations de modèles. Une transformation génère un modèle dans un langage cible à partir d'un modèle dans un langage source. Pour concevoir une transformation, les développeurs doivent avoir une bonne connaissance des méta-modèles concernés ainsi que des langages de transformation, ce qui rend cette tâche difficile. Dans cette thèse, nous proposons d'assister l'écriture des transformations et plus généralement de comprendre comment une transformation opère. Nous adhérons à l'approche de transformation de modèles par l'exemple qui propose de créer une transformation de modèles à partir d'exemples de transformation. Cela permet d'utiliser la syntaxe concrète définie pour les méta-modèles, et cela évite donc de requérir que les développeurs aient une bonne maîtrise des méta-modèles utilisés. Dans ce contexte, nous proposons deux contributions. La première consiste à définir une méthode pour générer des règles de transformation opérationnelles à partir d'exemples. Nous nous basons sur une approche qui utilise l'Analyse Relationnelle de Concepts (ARC) comme technique d'apprentissage pour obtenir des patrons de transformation à partir d'un appariement de type 1-1 entre les modèles. Nous développons une technique pour extraire des règles de transformation opérationnelles à partir de ces patrons. Ensuite, nous utilisons le langage et le moteur de règles JESS pour exécuter ces règles. Nous étudions aussi comment mieux apprendre des règles de transformations à partir d'exemples, en utilisant séparément chaque exemple ou en réunissant tous les exemples. La deuxième contribution consiste à récupérer les traces de transformation à partir d'exemples de transformation. Ces traces peuvent être utilisées par exemple pour localiser des erreurs durant l'exécution des programmes de transformation ou vérifier la couverture de tous les modèles d'entrée par une transformation. Dans notre contexte, nous supposons que ces traces vont servir pour un futur apprentissage des règles de transformation. Nous traitons tout d'abord le problème de récupération des traces avec des exemples provenant d'un programme de transformation. Nous proposons une approche basée sur une méta-heuristique multi-objectifs pour générer des traces sous forme d'appariement de type n-m entre des éléments de modèles. La fonction objectif s'appuie sur une similarité lexicale et structurelle entre ces éléments. Une extension de cette méthode est proposée pour traiter le problème plus général de l'appariement entre modèles. / Model Driven Engineering (MDE) considers models as first class artifacts. Each model conforms to another model, called its metamodel which defines its abstract syntax and its semantics.Various kinds of models are handled successively in an MDE development cycle. They are manipulated using, among others, programs called model transformations. A transformation takes as input a model in a source language and produces a model in a target language. The developers of a transformation must have a strong knowledge about the source and target metamodels which are involved and about the model transformation language. This makes the writing of the model transformation difficult.In this thesis, we address the problem of assisting the writing of a model transformation and more generally of understanding how a transformation operates.We adhere to the Model Transformation By example (MTBE) approach, which proposes to create a model transformation using examples of transformation. MTBE allows us to use the concrete syntaxes defined for the metamodels. Hence, the developers do not need in-depth knowledge about the metamodels. In this context, our thesis proposes two contributions.As a first contribution, we define a method to generate operational transformation rules from transformation examples. We extend a previous approach which uses Relational Concept Analysis as a learning technique for obtaining transformation patterns from 1-1 mapping between models. We develop a technique for extracting relevant transformation rules from these transformation patterns and we use JESS language and engine to make the rules executable. We also study how we better learn transformation rules from examples, using transformation examples separately or by gathering all the examples.The second contribution consists in recovering transformation traces from transformation examples. This trace recovery is useful for several purposes as locating bugs during the execution of transformation programs, or checking the coverage of all input models by a transformation. In our context, we expect also that this trace will provide data for a future model transformation learning technique. We first address the trace recovery problem with examples coming from a transformation program. We propose an approach, based on a multi-objective meta-heuristic, to generate the textit{many-to-many} mapping between model constructs which correspond to a trace. The fitness functions rely on the lexical and structure similarity between the constructs. We also refine the approach to apply it to the more general problem of model matching.

No description available.

No description available.

Made available in DSpace on 2016-08-17T14:53:26Z (GMT). No. of bitstreams: 1 Dissertacao Marcus Vinicius Ribeiro.pdf: 4694533 bytes, checksum: b84a4bad63b098d054781131cfb9bc26 (MD5) Previous issue date: 2014-02-24 / Model Driven Engineering (MDE) aims to make face to the development, maintenance and evolution of complex software systems, focusing in models and model transformations. This approach can be applied in other domains such as database schema integration. In this research work, we propose a framework to integrate database schema in the MDE context. Metamodels for defining database model, database model matching, database model merging, and integrated database model are proposed in order to support our framework. An algorithm for database model matching and an algorithm for database model merging are presented. We present also, a prototype that extends the MT4MDE and SAMT4MDE tools in order to demonstrate the implementation of our proposed framework, metodology, and algorithms. An illustrative example helps to understand our proposed framework. / A Engenharia Dirigida por Modelos (MDE) fornece suporte para o gerenciamento da complexidade de desenvolvimento, manutenção e evolução de software, através da criação e transformação de modelos. Esta abordagem pode ser utilizada em outros domínios também complexos como a integração de esquemas de base de dados. Neste trabalho de pesquisa, propomos uma metodologia para integrar schema de base de dados no contexto da MDE. Metamodelos para definição de database model, database model matching, database model merging, integrated database model são propostos com a finalidade de apoiar a metodologia. Um algoritmo para database model matching e um algoritmo para database model merging são apresentados. Apresentamos ainda, um protótipo que adapta e estende as ferramentas MT4MDE e SAMT4MDE a fim de demonstrar a implementação do framework, metodologia e algoritmos propostos. Um exemplo ilustrativo ajuda a melhor entender a metodologia apresentada, servindo para explicar os metamodelos e algoritmos propostos neste trabalho. Uma breve avaliação do framework e diretrizes futuras sobre este trabalho são apresentadas.

Autonomous vehicles have gained much importance over the last decade owing to their promising capabilities like improvement in overall traffic flow, reduction in pollution and elimination of human errors. However, when it comes to long-distance transportation or working in complex isolated environments like mines, various factors such as safety, fuel efficiency, transportation cost, robustness, and accuracy become very critical. This thesis, developed at the Connected and Autonomous Systems department of Scania AB in association with KTH, focuses on addressing the issues related to fuel efficiency, robustness and accuracy of an autonomous heavy-duty truck used for mining applications. First, in order to improve the state prediction capabilities of the simulation model, a comparative analysis of two dynamic bicycle models was performed. The first model used the empirical PAC2002 Magic Formula (MF) tyre model to generate the tyre forces, and the latter used a piece-wise Linear approximation of the former. On top of that, in order to account for the nonlinearities and time delays in the lateral direction, the steering dynamic equations were empirically derived and cascaded to the vehicle model. The fidelity of these models was tested against real experimental logs, and the best vehicle model was selected by striking a balance between accuracy and computational efficiency. The Dynamic bicycle model with piece-wise Linear approximation of tyre forces proved to tick-all-the-boxes by providing accurate state predictions within the acceptable error range and handling lateral accelerations up to 4 m/s2. Also, this model proved to be six times more computationally efficient than the industry-standard PAC2002 tyre model. Furthermore, in order to ensure smooth and accurate driving, several Model Predictive Control (MPC) formulations were tested on clothoid-based Single Lane Change (SLC), Double Lane Change (DLC) and Truncated Slalom trajectories with added disturbances in the initial position, heading and velocities. A linear time-varying Spatial error MPC is proposed, which provides a link between spatial-domain and time-domain analysis. This proposed controller proved to be a perfect balance between fuel efficiency which was achieved by minimising braking and acceleration sequences and offset-free tracking along with ensuring that the truck reached its destination within the stipulated time irrespective of the added disturbances. Lastly, a comparative analysis between various Prediction-Simulation model pairs was made, and the best pair was selected in terms of its robustness to parameter changes, simplicity, computational efficiency and accuracy. / Under det senaste årtiondet har utveckling av autonoma fordon blivit allt viktigare på grund av de stora möjligheterna till förbättringar av trafikflöden, minskade utsläpp av föroreningar och eliminering av mänskliga fel. När det gäller långdistanstransporter eller komplexa isolerade miljöer så som gruvor blir faktorer som bränsleeffektivitet, transportkostnad, robusthet och noggrannhet mycket viktiga. Detta examensarbete utvecklat vid avdelningen Connected and Autonomous Systems på Scania i samarbete med KTH fokuserar på frågor gällande bränsleeffektivitet, robusthet och exakthet hos en autonom tung lastbil i gruvmiljö. För att förbättra simuleringsmodellens tillståndsprediktioner, genomfördes en jämförande analys av två dynamiska fordonsmodeller. Den första modellen använde den empiriska däckmodellen PAC2002 Magic Formula (MF) för att approximera däckkrafterna, och den andra använde en stegvis linjär approximation av samma däckmodell. För att ta hänsyn till ickelinjäriteter och laterala tidsfördröjningar inkluderades empiriskt identifierade styrdynamiksekvationer i fordonsmodellen. Modellerna verifierades mot verkliga mätdata från fordon. Den bästa fordonsmodellen valdes genom att hitta en balans mellan noggrannhet och beräkningseffektivitet. Den Dynamiska fordonsmodellen med stegvis linjär approximation av däckkrafter visade goda resultat genom att ge noggranna tillståndsprediktioner inom det acceptabla felområdet och hantera sidoacceleration upp till 4 m/s2 . Den här modellen visade sig också vara sex gånger effektivare än PAC2002-däckmodellen. v För att säkerställa mjuk och korrekt körning testades flera MPC varianter på klotoidbaserade trajektorier av filbyte SLC, dubbelt filbyte DLC och slalom. Störningar i position, riktining och hastighet lades till startpositionen. En MPC med straff på rumslig avvikelse föreslås, vilket ger en länk mellan rumsdomän och tidsdomän. Den föreslagna regleringen visade sig vara en perfekt balans mellan bränsleeffektivitet, genom att minimering av broms- och accelerationssekvenser, och felminimering samtidigt som lastbilen nådde sin destination inom den föreskrivna tiden oberoende av de extra störningarna. Slutligen gjordes en jämförande analys mellan olika kombinationer av simulerings- och prediktionsmodell och den bästa kombinationen valdes med avseende på dess robusthet mot parameterändringar, enkelhet, beräkningseffektivitet och noggrannhet.