Ces travaux de thèse portent sur l'intégration hétérogène des fonctions de commande pour des transistors de puissance verticaux à grille isolée. Ce travail a consisté en la conception des fonctions de commande pour un transistor de type MOSFET en technologie CMOS planar, puis en la conception du composant de puissance lui-même, incluant des fonctions spécifiques pour l'auto-alimentation de sa commande. Le deuxième aspect de ce travail est l'étude et la réalisation technologique de métallisations épaisses en surface de puces silicium pour l'hybridation en chip on chip de la partie commande et de la partie puissance. Ce mémoire de thèse comporte trois chapitres équivalents : études théoriques et présentation des concepts, conception et validation expérimentale de la partie commande puis conception de la partie puissance et développements technologiques. Les champs d'application de ces travaux sont variés car ils couvrent un large domaine de l'électronique de puissance (convertisseurs hybrides). / The work presented in this PhD manuscript deals with heterogeneous integration of the control functions for driving a vertical power MOSFET. This work consisted in the design of the driving functions for a vertical power MOSFET in a planar CMOS technology, then in the design of the power transistor itself, including specific functions for the self powering of its driving circuitry. The second aspect of this work was the technological study and realization of thick metallization at the surface of silicon dies for the chip on chip assembly of the control and the power parts. This manuscript is composed of three equal chapters: the theoretical study and concepts presentation, the design and experimental validation of the driver part, and the design of the power part and the technological realizations. The field of application of this work covers a large spectrum as it concerns every hybrid converter

abstract: Point of Load (PoL) converters are important components to the power distribution system in computer power supplies as well as automotive, space, nuclear, and medical electronics. These converters often require high output current capability, low form factor, and high conversion ratios (step-down) without sacrificing converter efficiency. This work presents hybrid silicon/gallium nitride (CMOS/GaN) power converter architectures as a solution for high-current, small form-factor PoL converters. The presented topologies use discrete GaN power devices and CMOS integrated drivers and controller loop. The presented power converters operate in the tens of MHz range to reduce the form factor by reducing the size of the off-chip passive inductor and capacitor. Higher conversion ratio is achieved through a fast control loop and the use of GaN power devices that exhibit low parasitic gate capacitance and minimize pulse swallowing. This work compares three discrete buck power converter architectures: single-stage, multi-phase with 2 phases, and stacked-interleaved, using components-off-the-shelf (COTS). Each of the implemented power converters achieves over 80% peak efficiency with switching speeds up-to 10MHz for high conversion ratio from 24V input to 5V output and maximum load current of 10A. The performance of the three architectures is compared in open loop and closed loop configurations with respect to efficiency, output voltage ripple, and power stage form factor. Additionally, this work presents an integrated CMOS gate driver solution in CMOS 0.35um technology. The CMOS integrated circuit (IC) includes the gate driver and the closed loop controller for directly driving a single-stage GaN architecture. The designed IC efficiently drives the GaN devices up to 20MHz switching speeds. The presented controller technique uses voltage mode control with an innovative cascode driver architecture to allow a 3.3V CMOS devices to effectively drive GaN devices that require 5V gate signal swing. Furthermore, the designed power converter is expected to operate under 400MRad of total dose, thus enabling its use in high-radiation environments for the large hadron collider at CERN and nuclear facilities. / Dissertation/Thesis / Masters Thesis Electrical Engineering 2018

This thesis presents System Identification based On-Line Health Monitoring to analyse the dynamic behaviour of the Switch-Mode Power Converter (SMPC), detect, and diagnose anomalies in passive electronic components. The anomaly detection in this research is determined by examining the change in passive component values due to degradation. Degradation, which is a long-term process, however, is characterised by inserting different component values in the power converter. The novel health-monitoring capability enables accurate detection of passive electronic components despite component variations and uncertainties and is valid for different topologies of the switch-mode power converter. The need for a novel on-line health-monitoring capability is driven by the need to improve unscheduled in-service, logistics, and engineering costs, including the requirement of Integrated Vehicle Health Management (IVHM) for electronic systems and components. The detection and diagnosis of degradations and failures within power converters is of great importance for aircraft electronic manufacturers, such as Thales, where component failures result in equipment downtime and large maintenance costs. The fact that existing techniques, including built-in-self test, use of dedicated sensors, physics-of-failure, and data-driven based health-monitoring, have yet to deliver extensive application in IVHM, provides the motivation for this research ... [cont.].

L’intégration croissante de la Génération Distribuée basée sur des sources d’énergies renouvelables présente de nouveaux défis pour le réseau électrique centralisé actuel. Le micro-réseau est une des alternatives envisagées afin d’augmenter le taux de pénétration d’énergie renouvelable et d’améliorer la qualité de l’énergie. La stabilité d’un micro-réseau est fortement sensible aux variations de puissance venant des sources d’énergie ou des charges. Dans ce contexte, un système de stockage d’énergie joue un rôle essentiel et doit satisfaire deux conditions : disposer d’une capacité de stockage importante pour adapter la production à la demande et être capable de fournir rapidement une puissance instantanée suffisante pour pallier les problèmes de qualité d’énergie. L’objectif principal de cette thèse est de concevoir et valider expérimentalement un système de conversion de puissance et l’algorithme de contrôle associé pour la gestion du stockage dans un micro-réseau afin de satisfaire les deux conditions. Suite à l’analyse de différentes technologies de stockage, on peut conclure qu’il n’y a pas de système de stockage capable de satisfaire les conditions d’énergie et de puissance en même temps. Par conséquent, l’association d’un supercondensateur et d’une batterie Redox au Vanadium dans un Système de Stockage Hybride est utilisée pour satisfaire les conditions mentionnées. Le travail de recherche est axé sur la gestion du flux d’énergie et de puissance du Système de Stockage Hybride proposé à l’aide d’un système de conversion de puissance innovant et son algorithme de commande. Un convertisseur multi-niveaux 3LNPC a été choisi pour commander en même temps les deux systèmes de stockage, en raison de faibles pertes de puissance et de distorsions harmoniques réduites en comparaison avec d’autres topologies existantes. Un algorithme de commande capable d’exploiter les limites de fonctionnement du convertisseur sur toute sa plage de fonctionnement a été conçu afin de satisfaire de manière optimale les critères spécifiés. Le fonctionnement du système de conversion et la stratégie de commande proposée ont été validés d’abord en simulation et ensuite expérimentalement en utilisant le micro-réseau installé à l’ESTIA. / The increasing penetration of Distributed Generation systems based on Renewable Energy Sources is introducing new challenges in the current centralised electric grid. The microgrid is one of the alternatives that is being analysed in order to increase the penetration level of renewable energy sources in electrical grids and improve the power quality. The microgrid stability is highly sensitive to power variations coming from the energy sources or loads. In this context, an energy storage system is essential and it must satisfy two criteria: to have a high storage capacity to adapt the generation to the demand and to be able to supply fast power variations to overcome the power quality problems that may arise. The main objective of this thesis has been to design a power conversion system and the associated control algorithm for a storage system management in order to satisfy the defined requirements, as well as to experimentally validate the proposed solution. After an analysis of different storage system technologies, it can be concluded that there is not any storage system capable of offering the energy and power requirements at the same time. Consequently, the association of a SuperCapacitor bank and a Vanadium Redox Battery is used to satisfy the mentioned requirements. This thesis has been focused on the power and energy flow management of the proposed Hybrid Energy Storage System through an innovative power conversion system and its control method. A Three-Level Neutral Point Clamped converter has been used to control at the same time the two storage systems, due to the reduced power losses and harmonic distortion compared to other existing topologies. A control algorithm that uses the operational limits of the converter in its entire operation range has been designed in order to allow selecting the best operation point according to the specified criteria. The operation of the power conversion system and the proposed control method have been first validated in simulations and then experimentally using the microgrid installed in ESTIA.

The trend of future wind farms moving further offshore requires much higher reliability for each wind turbine in order to reduce maintenance cost. The drive-train system and power electronic converter system have been identified as critical sub-assemblies that are subject to higher failure rates than the other sub-assemblies in a wind turbine. Modern condition monitoring techniques may help schedule the maintenance and reduce downtime. However, when it comes to offshore wind turbines, it is more crucial to reduce the failure rates (or reduce the stresses) for the wind turbines during operation since the harsh weather and a frequently inaccessible environment will dramatically reduce their availability once a failure happens. This research examines the mechanical, electrical and thermal stresses in the sub-assemblies of a doubly-fed induction generator (DFIG) wind turbine and how to reduce them by improved control strategies. The DFIG control system (the rotor-side and the grid-side converter control) as well as the wind turbine control system are well established. The interactions of these control systems have been investigated. This research examines several further strategies to reduce the mechanical and electrical stresses. The control system's coordination with the protection schemes (crowbar and dc-chopper) during a grid fault is presented as well. An electro-thermal model of the power converter has been developed to integrate with the DFIG wind turbine model, for the evaluation of the thermal stresses under different operating states and control schemes. The main contributions of this thesis are twofold. A first contribution is made by providing all the control loops with well-tuned controllers in a more integrated methodology. The dynamics of these controllers are determined from their mathematical models to minimize the interference between different control-loops and also to reduce the electrical transients. This thesis proposes a coordination strategy for the damping control, pitch control and crowbar protection which significantly reduces the mechanical oscillations. On the other hand, an integrated model of the wind turbine and converter electro-thermal system is established that can illustrate the performance integration with different control strategies.

This thesis report details the design and construction of non-isolated DC-DC converters to create a Multiple Input Single Output (MISO) converter for combining multiple renewable energy sources into one single output. This MISO uses the four-switch buck-boost topology to output a single 48V from multiple nominal 24V inputs. The MISO converter implements a modular approach to deliver 1800W output power. Each module in the MISO is rated at 600W and they share the output power equally. Hardware results show that the converter produces 1800W of output power from three sources with 96.4% efficiency. Each module also demonstrates equal sharing feature of the MISO converter.

The dissertation presents a generalized control structure for two-stage power converters operated in a renewable energy power system for smart grid and micro grid systems. The generalized control structure is based on the two-loop average-mode-control technique, and created by reconstructing the conventional control structure and feedback configuration. It is broadly used for both dc-dc and dc-ac power conversion based on the two-stage converter architecture, while offering several functionalities required for renewable energy power systems. The generalized control structure improves the performance and reliability of renewable energy power systems with multiple functionalities required for consistent and reliable distributed power sources in the applications of the smart grid and micro grid system. The dissertation also presents a new modeling approach based on a modification of the subsystem-integration approach. The approach provides continuous-time small-signal models for all of two-stage power converters in a unified way. As a result, a modeling procedure is significantly reduced by treating a two-stage power converter as a single-stage with current sinking or sourcing. The difficulty of linearization caused by time-varying state variables is avoided with the use of the quasi-steady state concept. The generalized control structure and modeling approach are demonstrated using the two-stage dc-dc and dc-ac power conversion systems. A battery energy storage system with a thermoelectric source and a grid-connected power system with a photovoltaic source are examined. The large-signal averaged model and small-signal model are developed for the two demonstrated examples, respectively. Based on the modeling results, the control loops are designed by using frequency domain analysis. Various simulations and experimental tests are carried out to verify the compensator designs and to evaluate the generalized control structure performance. From the simulation and experimental results, it is clearly seen that the generalized control structure improves the performance of a battery energy storage system due to the unified control concept. The unified control concept eliminates transient over-voltage or over-current, extra energy losses, power quality issues, and complicated decision processes for multiple-mode control. It is also seen that the generalized control structure improves the performance of a single-phase grid-connected system through increased voltage control loop bandwidth of the active ripple current reduction scheme. As a result of the increased loop bandwidth, the transient overshoot or undershoot of the dc-link voltage are significantly reduced during dynamic load changes. / Ph. D.

A novel multilevel converter that is especially suited for high speed multi-megawatt switched reluctance motor drives operating at the medium voltage level is presented. The drive is capable of variable speed, four-quadrant operation. Each phase leg of the converter contains an arbitrary number of cascaded cells connected in series with the phase winding. Each cell contains a half-bridge chopper connected to a capacitor. The converter is named the cascaded chopper cell converter. The modular nature of the converter with the ability to add redundant cells makes it very reliable, which is a key requirement for medium voltage drive applications. A comprehensive control algorithm that overcomes the challenges of balancing and controlling cell capacitor voltages is also proposed. A suitable startup algorithm to limit startup current and switching losses, as well as ensure that cell capacitor voltages remain controlled at startup, is suggested. Details of the drive design such as component sizing and control parameter selection are also discussed. A detailed simulation model is developed and explained, and simulation results are provided for primary validation. Operation with standard current and speed control is first simulated. Then a scheme that gives way to a controller that operates the drive in single-pulse mode is developed and presented. This single-pulse control scheme controls the turn-on and turn-off angles, as well as the energization voltage level, in order to obtain high efficiency. Practical considerations related to the drive such as reliability, efficiency, and cost considerations are also discussed. Finally, a detailed comparison of the proposed converter to another competing converter is performed. Besides its scalability to high voltages and powers, the reliability and efficiency of the proposed converter makes it also a candidate for sub-megawatt applications requiring minimum downtime, or any application where high efficiency or improved performance is required. A small part of this work is also dedicated to brushless dc machines. Control methods for a new converter for brushless dc machines are proposed and verified via simulation. The main advantage of this converter with the proposed control is that it allows exact control of torque or speed up to twice the rated speed, without resorting to current phase advancing or other flux-weakening techniques. / Ph. D.

O presente trabalho apresenta o estudo, o equacionamento, a simulação e uma metodologia de projeto para um conversor CC/CC do tipo Half-Bridge, com transformador de potência alimentado por transistores bipolares em semi-ponte, com circuito de acionamento de base proporcional isolado. O circuito possui um controle de sincronismo aplicado ao transformador de pulso responsável pelo controle das chaves semicondutoras do conversor. A metodologia de projeto do circuito de acionamento de base proporcional isolado é discutida e alguns resultados são mostrados. Além disso, pretende-se empregar esse tipo de conversor em veículos aeroespaciais devido a robustez no circuito de acionamento de base, em específico na Unidade de Suprimento de Energia do satélite universitário ITASAT. Utilizando-se o diagrama esquemático do conversor proposto é feito o equacionamento matemático, a fim de se obter a frequência de oscilação livre, ou natural, a partir do qual é proposta uma metodologia de projeto. Ao final do trabalho são comparados os resultados experimentais com a simulação, com o objetivo de melhor avaliar a metodologia proposta. / This work presents the study, equation development, simulation and a design methodology for a Half-Bridge DC/DC converter, with power transformer fed by bipolar transistors in a half-bridge configuration, with proportional base drive circuit. The circuit has a synchronism control applied to the pulse transformer responsible for controlling the semiconductor switches of the converter. The design methodology of the proportional base drive is discussed and some results are shown. In addition, this type of converter is intended to be used in aerospace vehicles due to the increased robustness of the base drive circuit, for instance, the Power Supply Unit of ITASAT Satellite. A mathematical modeling is done in order to obtain the oscillation frequency (natural frequency) of the converter, from which a design methodology is proposed. At the end of the study the experimental are compared with the simulation, in order to evaluate the proposed methodology.

Cette thèse s’intéresse à la commande et l’observation d’une classe de systèmes linéaires à commutations (SLC). La classe considérée regroupe les systèmes pouvant être représentés par un modèle Hamiltonien à ports. Récemment, plusieurs travaux ont utilisé la théorie des systèmes dynamiques hybrides pour traiter les problèmes de stabilité, de commandabilité et d’observabilité des systèmes linéaires à commutations. Cependant, certains verrous scientifiques demeurent et nécessitent d’être levés tels que la synthèse d’observateurs pour des SLC présentant des modes de fonctionnement inobservables ou la commande hybride de systèmes possédant un nombre réduit d’entrées de commutations et un nombre élevé de variables d’état à contrôler. Dans ce travail, nous nous intéressons à la synthèse d’observateurs s’appuyant sur la modélisation moyenne et la modélisation hybride de SLC ayant une topologie Hamiltonienne à ports particulière. Ce formalisme possède les outils nécessaires pour établir des preuves de stabilité des erreurs d’observation. Dans un premier temps, nous proposons un observateur non linéaire reposant sur le modèle moyen de la classe des SLC considérée. Ensuite, nous traitons le problème de synthèse d’un observateur hybride où nous proposons un observateur commuté prenant en compte les modes de fonctionnement inobservables. Le problème de la commande des SLC est abordé par la suite. Au départ, la théorie de Lyapunov est utilisée pour proposer deux lois de commandes : La première est synthétisée à partir du modèle moyen et la deuxième exploite le modèle hybride. Une commande optimale hybride est élaborée en utilisant le principe du maximum de Pontryagin et une approche utilisant la recherche d’arcs singuliers. Finalement, une commande prédictive hybride est établie à partir d’un modèle discrétisé du système. Des résultats de simulation et une mise en œuvre expérimentale sur un convertisseur DC-DC SEPIC sont donnés pour montrer l’efficacité des méthodes proposées. L’étude d’un tel circuit est motivée par sa topologie particulière qui contient à la fois un mode de fonctionnement observable et un mode de fonctionnement inobservable. En outre, il possède une seule entrée de commutations et quatre variables d’état ce qui lui vaut la réputation être difficile à commander. / This thesis is dedicated to the control and the observation of a class of switched linear systems (SLS). This class contains the systems that can be represented by a port-Hamiltonian model. A lot of works have been studied SLS for several years using an average modeling approach. Recently, various works have shown that hybrid system theory allows to cope with stabilization, controllability, and observability problems of switched linear systems. However, several problems are still open and need more development such as the design of hybrid observers for SLS that have unobservable modes or the control of systems with reduced number of switching inputs and numerous variable states to control. In this work, we are interested in the design of state observers for a particular class of SLS using both the average and the hybrid port-Hamiltonian models. This formalism has the necessary tools to study and establish the stability of the observation errors. At the beginning, a nonlinear observer based on the average modeling is proposed. Next, a hybrid observer is designed for switched linear systems. This observer takes into account the unobservable operating modes of the system. The second point of our work concerns the design of control laws for the considered class of SLS. At first, two Lyapunov-based control laws have been established using either an average model or a hybrid model of the system. A hybrid optimal control based on the maximum principle of Pontryagin and the computation of singular arcs has been also proposed. Finally, a hybrid predictive control based on a discrete model of the system is synthesized. Simulation results and an experimental implementation on a SEPIC converter are given to show the efficiency of the proposed methods. Our motivation to study such a converter is mainly due to its particular topology that includes observable and unobservable subsystems. It is also known to be difficult to be controlled because only one switching input is used to control four state variables.