Contribution au développement d'un système de controle avancé pour les systèmes éolien-diesel autonomes

The contribution of the thesis consists in the development of advanced models for the dynamic behaviour of a wind-diesel system as well as for the short-term prediction of the power output of a wind farm. In order to simulate the transient behaviour of an autonomous system, a generic model has been developed that incorporates models for all the power system components. A methodology has been developed for the robust identification of the unknown power system parameters. Following this development, the next step was to develop forecasting models based on recurrent high order neural networks and on fuzzy logic for the power output of a wind farm. The developed models on the wind power forecasting problem were found to outperform existing approaches in the literature. The consideration of the developed models in the control system contributes to improve the power system management and to increase wind power penetration while maintaining a high level of security in the network.