Off-grid hybrid renewable energy systems are known as an attractive and sustainable solution for supplying clean electricity to autonomous consumers. Typically, this applies to the communities that are located in remote or islanded areas where it is not cost-effective to extend the grid facilities to these regions. In addition, the use of diesel generators for electricity supply in these remote locations are proven to be uneconomical due to the difficult terrain which translates into high fuel transportation costs. The use of renewable energy sources, coupling with the diesel generator allows for the diesel fuel to be offset. However, to date, a common design standard for the off-grid system has yet to be found and some challenges still exist while attempting to design a reliable system. These include the sizing of hybrid systems, coordination between the operation of dissimilar power generators and the fluctuating load demands, optimal utilisation of the renewable energy resources and identifying the underlying principles which reduce the reliability of the off-grid systems. In order to address these challenges, this research has first endeavoured into developing a sizing algorithm which particularly seeks the optimal size of the batteries and the diesel generator usage. The batteries and diesel generator function in filling the gap between the power generated from the renewable energy resources and the load demand. Thus, the load requirement is also an important factor in determining the cost-effectiveness of the overall system in the long run. A sensitivity analysis is carried out to provide a better understanding of the relationship between the assessed renewable energy resources, the load demand, the storage capacity and the diesel generator fuel usage. The thesis also presents the modelling, simulation and experimental work on the proposed hybrid wind-diesel-battery system. These are being implemented with a full-scale system and they are based on the off-the-shelf components. A novel algorithm to optimise the operation of a diesel generator is also proposed. The steady-state and dynamic analysis of the proposed system are presented, from both simulation and an experimental perspective. Three single-phase grid-forming inverters and a fixed speed wind turbine are used as a platform for case studies. The grid-forming inverters adopt droop control method which allows parallel operation of several grid-forming sources. Droop control-based inverters are known as independent and autonomous due to the elimination of intercommunication links among distributed converters. Moreover, the adopted fixed speed wind turbine employs a squirrel cage induction generator which is well known for its robustness, high reliability, simple operation and low maintenance. The results show a good correlation between the modelling, the experimental measurements, and the field tested results. The final stage of this research explores the effect of tower shadow on off-grid systems. Common tower designs for small wind turbine applications, which are the tubular and the lattice configurations, are considered in this work. They generate dissimilar tower shadow profiles due to the difference in structure. In this research, they are analytically modelled for a wind turbine which is being constructed as a downwind configuration. It is proven that tower shadow indeed brings negative consequence to the system, particularly its influence on battery lifetime within an off-grid system. This detrimental effect occurs when power generation closely matches the load demand. In this situation, small frequent charging and discharging cycles or the so called microcycles, take place. The battery lifetime reduction due to these microcycles has been quantified and it is proven that they are not negligible and should be taken into consideration while designing an off-grid hybrid system.

Dans le domaine aéronautique, l'optimisation du rendement énergétique global, la réduction des masses embarquées et la nécessité de répondre aux besoins énergétiques croissants conduisent à développer de nouvelles technologies et méthodes pour générer l'énergie électrique à bord, pour la distribuer, la convertir et la stocker. Dans cette thèse, des éléments de stockage de l'énergie électrique sont caractérisés dans l'optique d'être modélisés. Parmi les différents systèmes de stockage, présentés dans un état de l'art préliminaire, sont retenus les supercondensateurs et les accumulateurs électrochimiques Lithium-ion polymère, considérés respectivement comme des sources de puissance et d'énergie, à l'échelle de l'application. Ces moyens de stockage sont caractérisés par chronopotentiométrie à courant constant et par spectrométrie d'impédance électrochimique. Les essais sont éffectués dans des conditions expérimentales, définissant le domaine de validités des modèles, en cohérence avec les contraintes de l'application finale. Différents modèles sont alors développés en fonction de leur utilisation : des modèles simples, fonctionnels et suffisants pour la gestion globale d'énergie et des modèles dynamiques, comportementaux et nécessaires pour l'analyse de la qualité du réseau. Ils sont ensuite validés sur des profils de mission. Pour disposer d'un système de stockage performant et en adéquation avec les besoins énergétiques de l'aéronef, une méthode de dimensionnement est proposée, associant des composants de stockage complémentaires. Un gestion fréquentielle des sources est mise en oeuvre de manière à minimiser la masse du système de stockage. / In aeronautics, the optimization of the global energetic efficiency, the reduction of the embedded weight and the need to meet the growing energetic requirements lead to develop new technologies and methods to generate electrical energy, to distribute it, to convert it and to store it aboard. In this thesis, electrical energy storage systems are characterized with a view to be modeled. Among varied storage systems, presented in an introductory state of the art, ultracapacitors and Lithium-ion polymer secondary batteries are studied. These components are considered respectively as power and energy sources, in regards to the application scale. These storage systems are characterized by chronopotentiometry at constant current and by electrochemical impedance spectrometry. Tests are carried out in experimental conditions which define the validity area of modeling, in relation with the application constraints. Different models are developed according to their future use : simple models, which are functional and sufficient for the global energy management, and dynamics models, which are behavioral and necessary for the analysis of the network quality. Then, they validated thanks to mission profiles. Finally, to dispose of an efficient storage system that meets the energetic requirements of the aircraft, a sizing method is suggested by combining complementary storage systems. An energy management based on frequency approach is implemented in order to minimize the storage system weight.