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.

Les micro-drones sont des aéronefs sans pilotes de dimensions inférieures à un mètre et de poids inférieur à deux kilogrammes. Ils se distinguent des aéronefs classiques pour plusieurs raisons : un cycle de développement plus court, un coût plus faible, leur facilité d'opération et des configurations de véhicules spécifiques. L'ensemble de ces points attendent une réponse spécifique dans le développement des lois de commandes. Cette thèse s'y intéresse à travers deux problématiques : la commande d'une famille de drones quadrirotors et celle d'un drone hybride. Une famille de drones représente un même concept de véhicule décliné en plusieurs tailles dont on peut faire varier la charge utile ou son emplacement. Les lois de commandes doivent assurer un même niveau de performances malgré ses modifications. Un drone hybride se caractérise par sa capacité à réaliser du vol stationnaire et du vol d'avancement. Ces deux modes de vol ont chacun une dynamique de vol spécifique à laquelle les lois de commandes doivent s'adapter. Cette thèse présente la modélisation de quadrirotors et d'un drone hybride puis détaille une approche de commande adaptative indirecte qui répond aux problèmes introduits. La commande adaptative permet de garantir à l'aide d'un correcteur unique les performances de commande pour de multiples systèmes. Les méthodes d'estimation de paramètres et de synthèse linéaire à paramètres variants du schéma de commande sont décrites, puis, finalement, des résultats d'essais en vol montrent l'apport et les limites de cette approche. / Micro Air Vehicle are pilotless aircrafts with dimensions not exceeding one meter and a maximum weight of two kilograms. They are different from classical aircrafts for multiple reasons: a shorter development cycle, a cheaper development, their ease of operation and specific vehicle configurations. All these points expect a specific answer in the development of the control laws of the vehicles. This thesis considers this topic through two particular issues: the control of a family of quadrotors and the control of hybrid micro air vehicle. A family of quadrotor represents a single concept of vehicle but with various sizes, payloads and payload configurations. Control laws must guarantee the same level of performance despite all these modifications. A hybrid micro air vehicle is able to both hover like a helicopter and fly forward like a plane. These two flight modes have specific flight dynamics that the control laws must adapt to. This thesis first presents a model of quadrotors and hybrid micro air vehicle and then details an indirect adaptive control method to tackle both issues. Adaptive control should guarantee performance of multiple controlled systems with a single controller. The parameter estimation and linear parameter varying synthesis method of the adaptive control scheme are described and finally flight test results show the contributions and limits of the approach.

The quest to balance occupant comfort with energy efficiency is a key challenge in the field of heating systems, particularly for hydronic radiators. This study addresses this issue by investigating the integration of body heat predictions into a gain-scheduling controller for a hydronic radiator system. Although the benefits of gain-scheduling control strategies are acknowledged in HVAC systems, this exploration into the integration of body heat predictions in hydronic radiator systems presents a novel approach. A Linear Parameter-Varying (LPV) controller was employed and its impact on comfort, energy consumption, and robustness in the face of varying parameters such as the number of occupants, inaccuracies in body heat prediction, and set-point temperature changes was examined. This proposed controller was tested in a simulated house heating system made in Simulink. Findings indicated a substantial enhancement in comfort, especially under low-load scenarios. The controller demonstrated notable robustness against disturbances, highlighting the system’s reliability. Although energy consumption did not show significant reduction, the ability to maintain comfort levels without increasing energy use is a valuable contribution to sustainable heating practices. The results of this study extend our understanding of control strategies in hydronic radiator systems, providing a promising approach towards more comfortable, robust, and energy-efficient solutions. Further research should focus on improving the accuracy of body heat prediction algorithms and incorporating renewable energy sources for increased energy efficiency. In sum, this work represents a significant step towards a more balanced and sustainable future in the operation of hydronic radiator systems. / Denna studie utforskar möjligheten att balansera komfort och energieffektivitet i vattenburna elementsystem genom att integrera kroppsvärmeberäkningar i en gain-scheduling regleralgoritm. Vi presenterar en nyanserad metod som använder en Linjär Parameterberoende (LPV) reglerare. Denna reglerare anpassar sig till varierande parametrar som antal personer i rummet, osäkerheter i kroppsvärmeberäkningar och förändringar i inställd temperatur. Den föreslagna regleraren testades i ett simulerat husvärmesystem skapat i Simulink. Resultaten indikerade en betydande förbättring i komfort, särskilt under låglastscenarier. Regleraren uppvisade också anmärkningsvärd robusthet mot störningar, vilket understrykersystemets tillförlitlighet. Även om ingen signifikant minskning i energiförbrukning observerades, är förmågan att bibehålla komfortnivåer utan att öka energianvändningen ett värdefullt bidrag till hållbara uppvärmningsmetoder. Denna studie utökar vår förståelse för reglerstrategier i vattenburna elementsystem och erbjuder en lovande väg framåt mot mer komfortabla, robusta och energieffektiva lösningar. För framtida forskning bör fokus ligga på att förbättra noggrannheten i kroppsvärmeberäkningsalgoritmer och att integrera förnybara energikällor för ökad energieffektivitet. Sammantaget representerar detta arbete ett betydande steg mot en mer balanserad och hållbar framtid i drift av vattenburna elementsystem.