Species distribution maps can provide important information to focus conservation efforts and enable spatial management of human activities. Two sympatric marine predators, grey seals (Halichoerus grypus) and harbour seals (Phoca vitulina), have overlapping ranges but contrasting population dynamics around the UK; whilst grey seals have generally increased, harbour seals have shown significant regional declines. A robust analytical methodology was developed to produce maps of grey and harbour seal usage estimates with corresponding uncertainty, and scales of spatial partitioning between the species were found. Throughout their range, both grey and harbour seals spend the majority of their time within 50 km of the coast. The scalability of the analytical approach was enhanced and environmental information to enable spatial predictions was included. The resultant maps have been applied to inform consent and licensing of marine renewable developments of wind farms and tidal turbines. For harbour seals around Orkney, northern Scotland, distance from haul out, proportion of sand in seabed sediment, and annual mean power were important predictors of space-use. Utilising seal usage maps, a framework was produced to allow shipping noise, an important marine anthropogenic stressor, to be explicitly incorporated into spatial planning. Potentially sensitive areas were identified through quantifying risk of exposure of shipping traffic to marine species. Individual noise exposure was predicted with associated uncertainty in an area with varying rates of co-occurrence. Across the UK, spatial overlap was highest within 50 km of the coast, close to seal haul outs. Areas identified with high risk of exposure included 11 Special Areas of Conservation (from a possible 25). Risk to harbour seal populations was highest, affecting half of all SACs associated with the species. For 20 of 28 animals in the acoustic exposure study, 95% CI for M-weighted cumulative Sound Exposure Levels had upper bounds above levels known to induce Temporary Threshold Shift. Predictions of broadband received sound pressure levels were underestimated on average by 0.7 dB re 1μPa (± 3.3). An analytical methodology was derived to allow ecological maps to be quantitatively compared. The Structural Similarity (SSIM) index was enhanced to incorporate uncertainty from underlying spatial models, and a software algorithm was developed to correct for internal edge effects so that loss of spatial information from the map comparison was limited. The application of the approach was demonstrated using a case study of sperm whales (Physeter macrocephalus, Linneaus 1758) in the Mediterranean Sea to identify areas where local-scale differences in space-use between groups and singleton whales occurred. SSIM is applicable to a broad range of spatial ecological data, providing a novel tool for map comparison.

De nos jours, l’exploitation des énergies renouvelables afin de générer de l’électricité est en croissance soutenue puisqu’elles sont à ressource illimitée, gratuites et ne provoquent pas de déchets ou d’émissions polluantes. Dans cette thèse, on se propose d’étudier l’un de ces types d’énergie à savoir l’énergie issue des courants marins. Il s’agit plus particulièrement de s’intéresser à la commande tolérante aux défauts des systèmes de récupération de l’énergie des courants marins. Le potentiel de la production d'électricité à partir des courants marins est estimé à une production de 100 GW dans le monde. Cependant, ces chaînes de conversion d’énergie sont exposées et soumises à des contraintes fonctionnelles et environnementales importantes et sévères. Ces contraintes favorisent inévitablement la dégradation des performances des différents blocs fonctionnels de ces systèmes et l’accélération de leur processus de vieillissement, conduisant ainsi à l’apparition des défauts d’origines mécaniques et/ou électriques. Ainsi, la mise en place des techniques de commande tolérantes aux défauts de ces systèmes permettra d’améliorer la fiabilité, les performances et réduire les coûts relatifs au fonctionnement en mode dégradé et aux opérations de maintenance. Le but des travaux de cette thèse est l’étude, la modélisation et la simulation d’une chaîne de conversion hydrolienne à vitesse variable dans le cas sain et le cas d’un défaut (soit au niveau de la machine synchrone à aimants permanents (défaut de la désaimantation) ou au niveau du convertisseur statique (défaut d’un circuit ouvert d’un interrupteur). Il s’agira donc d’étudier les différentes commandes tolérantes aux défauts utilisées en cas d’un défaut au niveau de la génératrice ou au niveau de l’électronique de puissance associée. / Nowadays, the exploitation of renewable energies in order to generate electricity is growing steadily because they are unlimited resources, free and don’t cause waste or polluting emissions. In this context, it is proposed to study one of these types of energy, which is marine currents energy. In particular, we are interested in fault-tolerant control of tidal turbines. The potential for power generation from marine currents is estimated at 100GW in the world. However, tidal turbines are submitted to severe operational and environmental constraints. These constraints inevitably will lead to these systems performance degradation and the acceleration of their aging process, thus leading to the occurrence of mechanical and/or electrical faults. The implementation of fault-tolerant control techniques will improve the tidal turbines reliability, performance, and reduce costs relating to maintenance operations. The aim of this thesis is to study, model, and simulate a tidal turbine system in healthy and faulty conditions (either in the converter (switch open circuit) or in the permanent magnet synchronous generator (magnet failure). Various fault-tolerant control approaches are therefore evaluated and compared under these specific failure It will therefore be necessary to study the various fault-tolerant controls used in the event of a fault in the generator or in the associated power electronics.