Mathematical models of physical systems are pervasive in engineering. These models can be used to analyze properties of the system, to simulate the system, or synthesize controllers. However, many of these models are too complex or too large for standard analysis and synthesis methods to be applicable. Hence, there is a need to reduce the complexity of models. In this thesis, techniques for reducing complexity of large linear time-invariant (lti) state-space models and linear parameter-varying (lpv) models are presented. Additionally, a method for synthesizing controllers is also presented. The methods in this thesis all revolve around a system theoretical measure called the H2-norm, and the minimization of this norm using nonlinear optimization. Since the optimization problems rapidly grow large, significant effort is spent on understanding and exploiting the inherent structures available in the problems to reduce the computational complexity when performing the optimization. The first part of the thesis addresses the classical model-reduction problem of lti state-space models. Various H2 problems are formulated and solved using the proposed structure-exploiting nonlinear optimization technique. The standard problem formulation is extended to incorporate also frequency-weighted problems and norms defined on finite frequency intervals, both for continuous and discrete-time models. Additionally, a regularization-based method to account for uncertainty in data is explored. Several examples reveal that the method is highly competitive with alternative approaches. Techniques for finding lpv models from data, and reducing the complexity of lpv models are presented. The basic ideas introduced in the first part of the thesis are extended to the lpv case, once again covering a range of different setups. lpv models are commonly used for analysis and synthesis of controllers, but the efficiency of these methods depends highly on a particular algebraic structure in the lpv models. A method to account for and derive models suitable for controller synthesis is proposed. Many of the methods are thoroughly tested on a realistic modeling problem arising in the design and flight clearance of an Airbus aircraft model. Finally, output-feedback H2 controller synthesis for lpv models is addressed by generalizing the ideas and methods used for modeling. One of the ideas here is to skip the lpv modeling phase before creating the controller, and instead synthesize the controller directly from the data, which classically would have been used to generate a model to be used in the controller synthesis problem. The method specializes to standard output-feedback H2 controller synthesis in the lti case, and favorable comparisons with alternative state-of-the-art implementations are presented.

The market for electric vehicles (EV) is growing rapidly. The rise of EVs is most prominent for passenger vehicles, but trucks and busses are also quickly becoming electrified. Scania aims to be the leader of this transition. A central part of the EV is the Lithium-ion battery. In order to use the battery in the most efficient manner a Battery Management System (BMS) is needed. A key part of the BMS is a model that describes the battery as a system where the input is the current and the output is the terminal voltage. The dynamics of the battery is affected by external factors, called scheduling variables, that should be taken in to account in order to acquire an accurate model. This thesis aims to capture this behavior by the identification of a Linear Parameter Varying (LPV) model that has State of Charge (SOC) and temperature as scheduling variables. The LPV model was identified by first performing a set of local system identifications at varying levels of the scheduling variables. From this, a set of different LPV model structures were set up and then optimized with the use of datasets with a wider coverage of the scheduling variables. The results showed that there are clear advantages in using an LPV model compared to a traditional constant model, but that the robustness of the model largely is dependent on the choice of the data used for optimization.

Unmanned aircraft systems (UAS) often encounter turbulent fields that perturb the aircraft from its desired target trajectory, or in a manner that increases the load factor. The aircraft's fixed dihedral angle, providing passive roll-stiffness, is often selected based on lateral-directional stability requirements for the vehicle. A study to predict the effect of an active dihedral system on lateral-directional stability and vertical gust rejection capability was conducted to assess the performance and feasibility of the system. Traditionally, the dihedral location begins at the root to maintain wing structural requirements, however, the active dihedral system was also evaluated for dynamic stability and gust rejection performance at alternative dihedral breakpoint locations. Simulations were completed using linear parameter-varying (LPV) models, derived from traditional Newtonian aircraft dynamics and associated kinematic equations, to improve the modeling of the nonlinear active dihedral system. The stability of the LPV system was evaluated using Lyapunov stability theory applied to switched linear systems, assessing bounds of operation for the dihedral angle and flapping rate. An ideal feedback controller was developed using a linear–quadratic regulator (LQR) for both a discrete gust model and a continuous gust model, and a gain scheduled LQR controller was implemented to show the benefits of gain scheduling with a parameter varying state and input model. Finally, a cost analysis was conducted to investigate the real-world benefit of altering the dihedral breakpoint location. The effects of the active dihedral system on battery capacity and consumption efficiency were observed and compared with the gust rejection authority. / Master of Science / Unmanned aircraft systems (UAS) often encounter wind disturbances that perturb the aircraft from its desired target trajectory, or in a manner that increases the force encountered on the vehicle. The aircraft's fixed dihedral angle, providing stiffness to roll rotations, is often selected based on stability and control requirements for the vehicle. A study to predict the effect of a flapping wing (active dihedral) system on the stability, control, and wind gust rejection capability is completed to assess the performance and feasibility of such a system. Traditionally, the dihedral location begins at the root to maintain wing structural requirements, however, the active dihedral system was also evaluated for stability and wind gust rejection performance at alternative locations along the wing where the dihedral could begin, with intention of finding the best location. Simulations were completed using a varying set of simplified models, obtained from traditional aircraft mechanics, to improve the modeling of the true complex active dihedral system. The stability of the system was evaluated using various theories applied to the linear systems in attempt to define a bounded operating region for the dihedral angle and flapping motion. An ideal controller for the system was developed using ideas from well documented linear control theory for both a single wind gust and a continuous wind gust model. A controller that varies with vehicle flapping motion was implemented to show the benefits of scheduling the controller with a parameter varying state and input model. Finally, a cost analysis was conducted to investigate the real-world benefit of altering the dihedral starting location. The effects of the active dihedral system on battery capacity and consumption efficiency were observed and compared with the total gust rejection capability.

This dissertation studies the model reduction and distributed control problems for interconnected systems, i.e., systems that consist of multiple interacting agents/subsystems. The study of the analysis and synthesis problems for interconnected systems is motivated by the multiple applications that can benefit from the design and implementation of distributed controllers. These applications include automated highway systems and formation flight of unmanned aircraft systems. The systems of interest are modeled using arbitrary directed graphs, where the subsystems correspond to the nodes, and the interconnections between the subsystems are described using the directed edges. In addition to the states of the subsystems, the adopted frameworks also model the interconnections between the subsystems as spatial states. Each agent/subsystem is assumed to have its own actuating and sensing capabilities. These capabilities are leveraged in order to design a controller subsystem for each plant subsystem. In the distributed control paradigm, the controller subsystems interact over the same interconnection structure as the plant subsystems. The models assumed for the subsystems are linear time-varying or linear parameter-varying. Linear time-varying models are useful for describing nonlinear equations that are linearized about prespecified trajectories, and linear parameter-varying models allow for capturing the nonlinearities of the agents, while still being amenable to control using linear techniques. It is clear from the above description that the size of the model for an interconnected system increases with the number of subsystems and the complexity of the interconnection structure. This motivates the development of model reduction techniques to rigorously reduce the size of the given model. In particular, this dissertation presents structure-preserving techniques for model reduction, i.e., techniques that guarantee that the interpretation of each state is retained in the reduced order system. Namely, the sought reduced order system is an interconnected system formed by reduced order subsystems that are interconnected over the same interconnection structure as that of the full order system. Model reduction is important for reducing the computational complexity of the system analysis and control synthesis problems. In this dissertation, interior point methods are extensively used for solving the semidefinite programming problems that arise in analysis and synthesis. / Ph. D. / The work in this dissertation is motivated by the numerous applications in which multiple agents interact and cooperate to perform a coordinated task. Examples of such applications include automated highway systems and formation flight of unmanned aircraft systems. For instance, one can think of the hazardous conditions created by a fire in a building and the benefits of using multiple interacting multirotors to deal with this emergency situation and reduce the risks on humans. This dissertation develops mathematical tools for studying and dealing with these complex systems. Namely, it is shown how controllers can be designed to ensure that such systems perform in the desired way, and how the models that describe the systems of interest can be systematically simplified to facilitate performing the tasks of mathematical analysis and control design.

L’identification de systèmes est un sujet très utilisé à la fois dans le monde académique et industriel. Des nombreuses méthodes d’identification de systèmes invariants dans le temps existent dans la littérature et beaucoup d’algorithmes sont utilisés dans la modélisation pratique des systèmes. Ces outils offrent des résultats satisfaisants, mais ils ne sont pas capables de reproduire le caractère non linéaire présent dans les comportements des systèmes physiques. Ce besoin a conduit à l’apparition de la classe des systèmes linéaires à paramètres variants (LPV), capable de modéliser les aspects non linéaires des systèmes. Dans le cadre de cette thèse, différentes méthodes d’identification « classiques » ont été étudiées et modifiées pour prendre en compte le cas des modèles LPV.Dans un premier temps une étude sur les représentations et discrétisations des systèmes LPV a été réalisée. Ensuite, les méthodes à erreur de prédiction ont été étudiées et appliquées en vue d’identifier le comportement latéral d’un véhicule, en considérant la vitesse du véhicule comme paramètre variant. Les méthodes à erreur de prédiction ont été également appliquées afin de modéliser un convertisseur de puissance Buck, dont le comportement est sensible au changement de la résistance de charge, considérée comme paramètre variant. L’étude a été poursuivie avec la conception d’une loi de commande H_∞ de type LPV, appliquée au cas du convertisseur.Finalement, les méthodes des sous-espaces classiques ont été abordées et modifiées pour identifier les modèles LPV et appliquées au cas du comportement latéral d’un véhicule. / The identification system is a topic widely used both in the academic world and industry. Several methods of identification of time invariant systems exist in literature and many algorithms are used in practice for modeling real systems. These tools offer satisfactory results, but they are not able to reproduce the non-linearity occurring in the behavior of physical systems. The necessity of more has led to the occurrence of the class of linear systems parameter varying (LPV), able to model the nonlinear system aspects. In this thesis, different classical identification methods have been studied and their structures were modified, in order to take into account the LPV models. First, a study of representations and discretization of LPV models was performed. In the sequel, the prediction error methods have been studied and modified in order to take into account LPV models. This method was used to identify the lateral behavior of a vehicle, considering the speed of the vehicle as varying parameter. The prediction error method has also been applied to model a Buck converter, the behavior of which is sensitive to the changes of load resistance, the considered varying parameter. The study was continued with the design of a H_∞ LPV control law, applied to the converter. Finally, subspace methods were studied, modified for LPV models and applied to identify the lateral behavior of a vehicle.

No description available.

Tensegrity-membrane systems are a class of new bar-tendon-membrane systems. Such novel systems can be treated as extensions of tensegrity structures and are generally lightweight and deployable. These two major advantages enable tensegrity-membrane systems to become one of the most promising candidates for lightweight space structures and gossamer spacecraft. In this dissertation, modeling and control of tensegrity-membrane systems is studied. A systematic method is developed to determine the equilibrium conditions of general tensegrity-membrane systems. Equilibrium conditions can be simplified when the systems are in symmetric configurations. For one-stage symmetric systems, analytical equilibrium conditions can be determined. Three mathematical models are developed to study the dynamics of tensegrity-membrane systems. Two mathematical models are developed based on the nonlinear finite element method. The other model is a control-oriented model, which is suitable for control design. Numerical analysis is conducted using these three models to study the mechanical properties of tensegrity-membrane systems. Two control strategies are developed to regulate the deployment process of tensegrity-membrane systems. The first control strategy is to deploy the system by a nonlinear adaptive controller and use a linear H∞ controller for rapid system stabilization. The second control strategy is to regulate the dynamics of tensegrity-membrane systems using a linear parameter-varying (LPV) controller during system deployment. A gridding method is employed to discretize the system operational region in order to carry out the LPV control synthesis. / Ph. D.

O uso de etanol combustível no Brasil é hoje considerado o mais importante programa de combustível comercial renovável do mundo, sendo um potencial substituto aos derivados de petróleo. O aumento de rendimento fermentativo e a diminuição das perdas são objetivos de estudo em diversos centros de pesquisa, sendo o estudo da modelagem matemática e simulação do processo de grande importância para tal. A presente pesquisa apresenta como função identificar um modelo matemático para a linhagem isolada de Saccharomyces cerevisiae PE-2, de maneira a otimizar a maneira como é realizada a sua alimentação através de um controle H∞ por representação quasi-LPV. São realizados 9 ensaios de fermentação em 3 temperaturas distintas sob mesmas condições de concentração de substrato entrante. Após a finalização dos experimentos e análises, realiza-se a estimativa dos parâmetros componentes das equações diferenciais que modelam a cinética fermentativa, através de um algoritmo Quasi-Newton. De posse do modelo matemático, desenvolve-se um controle otimizado para a temperatura de 33ºC (temperatura usual de controle no processo industrial), considerando os parâmetros \"s\" e \"v\" variantes no tempo e os parâmetros x = 150 g/L e p = 70 g/L fixados, sendo valores médios obtidos durante o experimento. A utilização do controle desenvolvido possibilita um aumento de produtividade na faixa de 10% com relação a alimentação realizada em laboratório. Os resultados finais comprovam a eficiência do modelo matemático desenvolvido, comparado a outros estudos semelhantes, a influência da temperatura nos parâmetros cinéticos e a possibilidade de otimizar o processo através de um controle avançado do processo. / The use of ethanol in Brazil is considered the most important commercial renewable fuel program in the world, with a potential substitute for oil products. The increase in fermentation yield and losses reduction are objectives of study in various research centers, where the study of mathematical modeling and simulation of the process is of significant importance. This research presents as function to identify a mathematical model for the isolated strain of Saccharomyces cerevisiae PE-2, in order to optimize the way their substrate is fed, through a H∞ control based on quasi-LPV representation. Nine fermentation tests are performed at three different temperatures under the same conditions for incoming substrate concentration. After the experiments and analysis, it is carried out the estimation of parameters which are components of the differential equations that explain the fermentation kinetics, through a Quasi-Newton algorithm. With the mathematical model obtained, it is developed an optimal control for temperature 33°C (usual temperature control in the industrial process), considering the parameters \"s\" e \"v\" variyng in time and the parameters x = 150 g/L e p = 70 g/L set, which are average values obtained over the tests. The use of the control developed, applied to the flow variation, allows increasing productivity in 10% when compared with the flow performed in the tests conditions. The final results demonstrated the efficacy of the developed mathematical model, compared to other similar studies, the influence of temperature on the kinetic parameters and the possibility to optimize the process through an advanced process control.

Les systèmes décrits par les équations aux dérivées partielles, appartiennent à la classe des systèmes dynamiques impliquant des fonctions dépendantes de plusieurs variables, comme le temps et l'espace. Déjà fortement répandus pour la modélisation mathématique de phénomènes physiques et environnementaux, ces systèmes ont un rôle croissant dans les domaines de l'automatique. Cette expansion, provoquée par les avancées technologiques au niveau des capteurs facilitant l'acquisition de données et par les nouveaux enjeux environnementaux, incite au développement de nouvelles thématiques de recherche. L'une de ces thématiques, est l'étude des problèmes inverses et plus particulièrement l'identification paramétrique des équations aux dérivées partielles. Tout abord, une description détaillée des différentes classes de systèmes décrits par ces équations est présentée puis les problèmes d'identification qui leur sont associés sont soulevés. L'accent est mis sur l'estimation paramétrique des équations linéaires, homogènes ou non, et sur les équations linéaires à paramètres variant. Un point commun à ces problèmes d'identification réside dans le caractère bruité et échantillonné des mesures de la sortie. Pour ce faire, deux types d'outils principaux ont été élaborés. Certaines techniques de discrétisation spatio-temporelle ont été utilisées pour faire face au caractère échantillonné des données; les méthodes de variable instrumentale, pour traiter le problème lié à la présence de bruit de mesure. Les performances de ces méthodes ont été évaluées selon des protocoles de simulation numérique reproduisant des situations réalistes de phénomènes physique et environnementaux, comme la diffusion de polluant dans une rivière / A large variety of natural, industrial, and environmental systems involves phenomena that are continuous functions not only of time, but also of other independent variables, such as space coordinates. Typical examples are transportation phenomena of mass or energy, such as heat transmission and/or exchange, humidity diffusion or concentration distributions. These systems are intrinsically distributed parameter systems whose description usually requires the introduction of partial differential equations. There is a significant number of phenomena that can be simulated and explained by partial differential equations. Unfortunately all phenomena are not likely to be represented by a single equation. Also, it is necessary to model the largest possible number of behaviors to consider several classes of partial differential equations. The most common are linear equations, but the most representative are non-linear equations. The nonlinear equations can be formulated in many different ways, the interest in nonlinear equations with linear parameters varying is studied. The aim of the thesis is to develop new estimators to identify the systems described by these partial differential equations. These estimators must be adapted with the actual data obtained in experiments. It is therefore necessary to develop estimators that provide convergent estimates when one is in the presence of missing data and are robust to measurement noise. In this thesis, identification methods are proposed for partial differential equation parameter estimation. These methods involve the introduction of estimators based on the instrumental variable technique

La suspension joue un rôle central pour la dynamique verticale d’un véhicule automobile afin d’améliorer le confort des passagers et la tenue de route. Les travaux de recherche de cette thèse sont divisés en deux grandes parties. La première partie considère le problème de commande d’une suspension semi-active dont le défi principal est de prendre en compte les contraintes de dissipativité et de débattement maximum des amortisseurs. Celles-ci sont transformées en des contraintes sur la commande et l’état d‘un système linéaire. Deux approches sont alors proposées pour la synthèse de la commande de la suspension semi-active : la commande Linéaire à Paramètres Variants (LPV) avec prise en compte de la saturation et la Commande Prédictive à base de Modèle (MPC).La deuxième partie est consacrée à l’estimation de défaut actionneur et à la commande Tolérante à ce type de défauts, avec comme application majeure le système de suspension semi-active. On considère ici comme défaut une perte de puissance de l’amortisseur (par exemple une fuite de l’huile), qui est estimée en utilisant plusieurs approches fondées sur des observateurs d’état. Puis, en fonction de l’estimation du défaut, la commande en boucle fermée est reconfigurée afin de conserver des performances pour la dynamique verticale du véhicule. / Semi-active suspension system plays a key role in enhancing comfort and road holding of vertical dynamics in automotive vehicles. This PhD thesis research work, focused on that topic, is divided into two main parts. The first one considers the semi-active suspension control problem, the main challenge of which being to handle the dissipativity constraint and suspensions stroke limitation of semi-active dampers. These constraints are recast into input and state constraints in a linear state space representation. Thereby, the semi-active suspension control is designed in the framework of Linear Parameter Varying (LPV) approach with input constraints, and of Model Predictive Control (MPC) approach.The second part is devoted to Fault Estimation and Fault Tolerant Control (FTC) in case of actuator fault, and its application to Semi-Active suspension systems. The fault considered here is the loss of actuator's efficiency (due to an oil leakage of the damper for instance when a ), which is estimated using several observer-based approaches. Then, thanks to the fault information from the estimation step, an LPV/FTC fault scheduling control is designed to limit the vehicle performance deterioration.