Abstract:
This work focuses on the design of fault-tolerant control strategies to enhance the stability of wind turbines under actuator fault conditions. The main idea is to use the concept of observers redundancy to generate a residual set for an effective detection of faults in combination with a set of control schemes from which the adequate control strategy is automatically selected to maintain the operation of the system and ensure continuous and reliable energy supply during the occurrence of a fault. In our this work a bank of four observers is used for fault detection including Kalman filter, Sliding Mode observer, Luenberger observer and High gain observer. During no-fault condition (healthy mode), a Linear Quadratic Gaussian controller is used. However, when a fault occurs, the system automatically switches to a suitable controller depending on the fault nature and location. Generally, the controller known to have excellent properties of robustness against the parametric variations is used. due to it simple design. The system can be operated at low performance to keep the constant power deliverance. The thesis is divided in four main parts. First The general presentation of Wind conversion energy, In the second part Modeling and classical Control than we developed a new advanced control in third section, this parts are necessary groundwork is laid out concerning the dynamic wind turbine models ,This part form a basis for the last part Fault Diagnosis and Fault-Tolerant Control, where the fault diagnosis and fault-tolerant control schemes are described and applied to several fault scenario.
