Search

In this paper, an Hinfin control design methodology is presented in order to synthesize a full-order dynamic output feedback controller for an active power filter connected in parallel to a three-phase voltage distribution system. A shunt active filter is usually used to improve power quality by eliminating harmonic distortion generated by non-linear and switching loads. In order to be able to extract desired harmonics, the multivariable system is described by a linear coupled state space representation in a rotating two-phase framework. The originality of this paper is that axes decoupling is achieved by representing the system in a new complex framework in which the model is mathematically decoupled and therefore physical decoupling can be avoided. Moreover, the validity and efficiency of the proposed methodology are proved through numerical simulations.

In this paper, direct power control (DPC) of three-phase PWM rectifiers is presented, without line voltage sensors. The controller builds upon the ideas of the well known direct torque control (DTC) for induction motors. In our case, the active and reactive powers replace the torque and flux amplitude used as controlled output in DTC. The instantaneous active and reactive powers, directly controlled by selecting the optimum state of the converter, are used as the PWM control variables instead of the phase line currents being used. The controller ensures a good regulation of the output voltage, and guarantees the power factor close to one. The proposed controller was tested both in simulation and experimentally. The results show the excellent performance of the proposed system such as a good dynamic behavior, a good rejection of impact load disturbance.

The simulation and experimental study of new IP (Integrator Proportional) controlled DC bus voltage of three phase shunt active power filter (APF), to improve power quality by compensating harmonics and reactive power required by nonlinear load is proposed. The compensation process is based on sensing mains currents only, an approach different from conventional methods, which require harmonics or reactive volt-ampere values of the load. The not sinusoidal mains voltage form problem is resolved by using PLL system. PWM signal generation is based on hysteresis control comparators to obtain the switching signals. Various simulation and experimental results are presented under steady state and transient conditions.

The main objective of this work is to synthesize a robust static state feedback controller capable of controlling an active power filter connected in parallel to a three-phase voltage distribution system. A shunt active filter is used to improve power quality by eliminating harmonic distortion generated by non-linear loads. The open-loop system is described, in a complex frame, by a linear state space representation with complex-valued parameters. The synthesis of such controller is performed using H8 robust control based on a Linear Matrix Inequality (LMI) approach.

In this paper, a linear model of an active filter and its controller are proposed. The usual models are non linear due to the Park transform. The main interest of our work is the linearity of the model described as a state space representation. The originality of the model obtained is the complex valued state space representation. An experimental approach is proposed to perform controllers with complex valued parameters.

In this paper, harmonic suppression for power distribution systems with a parallel active filter is analysed. Together with harmonic compensation, the problem of LC resonances between the output filter of the three-phase inverter and the line inductances is dealt with. In our development, harmonic analysis and modelling system are presented before studying the means to control voltage at common point. In addition, a novel linear model of three-phase network with a parallel active filter is described as a complex state space representation. This linear model allows the use of usual tools for linear control theory. Theoretical analyses and a complete simulation of system are presented to confirm the validity of the proposed technique.

This paper presents a parallel active power filter which is intended for eliminating the unwanted harmonics. A novel linear model of a three-phase network is proposed and allows the use of tools developed for linear control theory. Moreover, with the same tools, uncertainties of physical parameters can be taken into account. Theoretical analysis, complete simulation of the system and experimental results are presented to confirm the validity of the proposed technique

Concerns the parameter identification of asynchronous machines. Extended Kalman filter is an optimal recursive estimation algorithm which allows combined state and parameter identification. In return, a good a priori knowledge on the parameters and on the stochastic properties of the noise disturbing both state and output are needed to initialize the nonlinear filter. Other methods use a linearized parametric model. Drawback of this approach is that the validity of this approximate model depends on assumptions about speed drive, which must remain constant. The estimation method described in this paper differs from other approaches in the following way: it is a recursive parameter identification method, based on a nonlinear continuous time model of the machine. A continuous time nonlinear dynamic model of the induction motor is presented. An output model where derivatives signals are not required is deduced from the reinitialized partial moment technique. After a description of the identification method, some considerations about the sensitivity functions are presented. Experimental results are also presented.

In this paper, a nonlinear programming algorithm is applied to the electrical parameter estimation of induction motors. A linear in the parameters output model of the system is obtained thanks to the reinitialized partial moments technique. This method is able to reconstruct the nonmeasured rotor currents. Simulation results show that this algorithm can perform an identification of the five electrical parameters of the motor.

Representing more than one-third of global electricity consumption, buildings undergo the most important sector capable of reducing greenhouse gas emissions and promote the share of Renewable Energy Sources (RES). The integrated RES and electric energy storage system in buildings can assist the energy transition toward a low-carbon electricity system while allowing end-energy consumers to benefit from clean energy. Despite its valuable advantages, this innovative distributed Building Microgrids (BM) topology requires significant changes in the current electric grid, which is highly dependent on grid energy policies and technology breakthroughs.The complexity of designing a robust Energy Management System (EMS) capable of managing all electric components inside the microgrid efficiently without harming the main grid stability is one of the greatest challenge in the development of BM. To mitigate the harmful effects of unpredictable grid actors, the concept of self-consumption has been increasingly adopted. Nonetheless, further technical-economic analysis is needed to optimally manage the energy storage systems to attain higher marks of self-consumption.Faceing these issues, the purpose of this doctoral thesis is to propose a complete framework for designing a building EMS for microgrids installed in buildings capable of maximising the self-consumption rate at minimum operating cost. Among all possible control architectures, the hierarchical structure has proved effective to handle conflicting goals that are not in the same timeframe. Hence, a Hierarchical Model Predictive (HMPC) control structure was adopted to address the uncertainties in the power imbalance as well as the trade-off between costs and compliance with the French grid code.Considering that buildings are not homogeneous and require solutions tailored to their specific conditions, the proposed controller was enhanced by two data-driven modules. The first data-driven algorithm is to handle inaccuracies in HMPC internal models. Without needing to tune any parameter, this algorithm can enhance the accuracy of the battery model up to three times and improve up to ten times the precision of the hydrogen storage model. This makes the building EMS more flexible and less dependent on pre-modelling steps.The second data-oriented algorithm determines autonomously adequate parameters to HMPC to relieve the trade-off between economic and energy aspects. Relying only on power imbalance data analysis and local measurements, the proposed hierarchical controller determines which energy storage device must run daily based on the estimation of the annual self-consumption rate and the annual microgrid operating cost. These estimations decrease microgrid expenditure because it avoids grid penalties regarding the requirements of annual self-consumption and reduces the degradation and maintenance of energy storage devices.The proposed EMS also demonstrated being capable of exploiting the potentials of shifting in time the charging of batteries of plug-in electric vehicles. The simulation confirmed that the proposed controller preferably charges electric vehicles’ batteries at periods of energy surplus and discharges them during periods of energy deficit, leading the building microgrid to reduce grid energy exchange. The results also showed that electric vehicle batteries' contribution depends on the size of the vehicle parking, their arrival and departure time, and the building’s net power imbalance profile. In conclusion, through simulations using the dataset of both public and residential buildings, the proposed hierarchical building EMS proved its effectiveness to handle different kinds of energy storage devices and foster the development of forthcoming building microgrids., Représentant plus d'un tiers de la consommation mondiale d'électricité, les bâtiments sont le secteur énergétique majeur pour promouvoir l’usage des énergies renouvelables. L'installation à la fois de sources d’énergie rénouvelable et d'un système de stockage d'énergie électrique dans les bâtiments peut favoriser la transition énergétique vers un système électrique à faible émission de carbone, tout en permettant aux consommateurs d'énergie finaux de bénéficier d'une énergie propre. Malgré tous ces avantages, cette topologie innovante et distribuée d’un Micro-réseau dédié au Bâtiment (MB)nécessite des changements importants dans le réseau actuel, qui dépend des politiques énergétiques et d’avancement technologiques.La conception d'un Système de Gestion de l'Energie (EMS) capable de gérer efficacement les composants électriques du micro-réseau sans menacer la stabilité du réseau principal est un obstacle au développement des MB. Pour atténuer les effets néfastes introduits par des acteurs d’énergie imprévisibles, le concept d'autoconsommation est de plus en plus adopté. Néanmoins, une analyse technico-économique plus approfondie est nécessaire pour piloter d’une manière optimaledes systèmes de stockage d'énergie afin d’atteindre des indices d'autoconsommation plus élevés.Face à ces enjeux, le but de ce doctorat est de proposer un EMS pour les micro-réseaux installés dans les bâtiments afin de maximiser leur taux d’autoconsommation à un coût d’exploitation minimum. Parmi les architectures de contrôle, la structure hiérarchique s'est avérée efficace pour gérer des objectifs contradictoires qui ne sont pas dans la même échelle de temps. Ainsi, une structure de contrôle Hiérarchique à Modèle Prédictif (HMPC) a été adoptée pour remédier aux incertitudes liées aux déséquilibres de puissance ainsi qu’établir un compromis entre la réduction du coût de fonctionnement et le respect du code de l’énergie français.Considérant que les bâtiments ne sont pas homogènes et nécessitent des solutions adaptées à leur besoin, le contrôleur proposé a été couplé à deux modules fonctionnant à base d’analyse de données. Le premier algorithme consiste à gérer les inexactitudes dans les modèles internes de l’HMPC. Sans avoir besoin de régler aucun paramètre, cet algorithme améliore la précision du modèle de batteries jusqu'à trois fois et augmente jusqu'à dix fois la précision du modèle de stockage d'hydrogène, réduisant ainsi la dépendance de l’EMS aux étapes de modélisation. Le deuxième algorithme détermine de manière autonome les paramètres de l’HMPC et facilite le compromis entre les aspects économiques et énergétiques. S'appuyant uniquement sur l'analyse des données de déséquilibre de puissance et des mesures, le contrôleur hiérarchique spécifie quel dispositif de stockage d'énergie doit fonctionner quotidiennement en fonction de l'estimation du taux d'autoconsommation et du coût de fonctionnement du micro-réseau. Ces estimations diminuent les dépenses annuelles du micro-réseau en évitant la pénalisation en ce qui concerne les exigences d'autoconsommation et en réduisant la dégradation et l'entretien des systèmes de stockage d'énergie.L’EMS proposé s'est également révélé capable de charger de préférence les batteries des véhicules électriques en période de surplus d’énergie et les décharger pendant les périodes de déficit pour réduire les échanges d’énergie avec le réseau principal. Les résultats ont aussi montré que la contribution des batteries de véhicules électriques dépend de la taille du parc de véhicules, de leur temps de connexion et du profil de déséquilibre de puissance. En conclusion, à travers les simulations utilisant le dimensionnement réel d'un bâtiment public et résidentiel, l’EMS hiérarchique s'est avéré efficace pour gérer de nombreux dispositifs de stockage d'énergie et contribuer à l’essor de micro-réseaux dédiés aux bâtiments à l’avenir.

This thesis in part of research thematic dedicated to photovoltaic solar energy production's systems. In fact, energy demands in the world are continuously increasing, the fossil fuels exploitation depletes limited sources and impact our environment. Solar photovoltaic energy represents one of the most promoted renewable energy sources, even if its drawbacks remain its installation’s high initial cost and its production’s variability. The main object of the presented work is to optimize the global efficiency of a stand-alone photovoltaic system. In that respect, this research focuses on the architecture of the conversion chain by integrating DC-DC conventional and quadratic boost converters and on the optimization of maximum power point tracking algorithms during fast changing weather conditions. These converters are modelling as switched affine systems to propose a switching rule that stabilize the operating point according to Lyapunav theory. Several simulation results using Matlab/SimulinkTM are given to declare the maximization of the extracted electrical power and the stability of the operating point. Experimental results recuperated on a test bench in the laboratory LIAS backed and confirmed the interest of the proposed approaches.; Cette thèse s'inscrit sur une thématique de recherche dédiée aux systèmes de production d'énergie solaire photovoltaïque. En effet, les besoins énergétiques mondiaux sont en croissance continue, l’exploitation et la consommation des énergies fossiles épuisent des ressources limitées et impactent notre environnement. L’énergie solaire photovoltaïque représente une source d'énergie renouvelable des plus prometteuse, même si ses inconvénients demeurent le coût élevé d'installation initiale et la variabilité de sa production. Les travaux présentés visent surtout à optimiser le rendement énergétique global d'un système solaire photovoltaïque non connecté au réseau. Ainsi, ce travail de recherche se focalise sur l’architecture de la chaîne de conversion avec l’intégration de convertisseurs élévateurs DC-DC conventionnel et quadratique et sur une optimisation des algorithmes de recherche du point de puissance maximale lors de variations des conditions climatiques. La modélisation et l’analyse de ces convertisseurs élévateurs sont conduites sous forme de systèmes à commutation afin de proposer une commande stabilisante aux points d’équilibre selon la définition de Lyapunov. Des résultats de simulation sous Matlab/SimulinkTM sont exposés, attestant la maximisation de la puissance électrique récupérée et la stabilité pour le point de fonctionnement. Des résultats expérimentaux, issus du banc expérimental du laboratoire LIAS, viennent corréler et confirmer l'intérêt des approches proposées.