Résumé:
Wind energy plays an increasingly crucial role in electricity production, requiring proactive
involvement of wind farms in managing the electrical grid. Additionally, diagnosing potential
faults in the wind energy chain and detecting failures are major priorities for both industry and
research. This thesis aims to enhance the energy quality of variable-speed wind turbines using a
doubly-fed induction generator (DFIG) by developing a comprehensive control method. The
primary objective is to regulate active and reactive power through field-oriented control (FOC) to
meet the requirements of the wind farm control system. To achieve this, three types of controllers
were investigated: proportional-integral (PI) controller, fuzzy logic controller (FLC), and neural
network-based controller (NNC). Simulations have shown the superiority of NNC in terms of
dynamic responsiveness and precise tracking of power reference values. Simultaneously, a
diagnostic method using fuzzy logic was devised to monitor and detect ITSC and open-phase
circuits in the stator windings of the DFIG within the wind system. This method solely utilizes
phase currents to detect and locate faults in real-time. Furthermore, to address converter failure
issues, a fault detection technique for pulse width modulation (PWM) inverters was studied. This
technique combines fuzzy logic and neural networks to accurately identify short-circuit and opencircuit faults in the wind generator inverter
