Voltage model observer based encoderless predictive control of DFIG for wind tubine applications

An encoderless predictive control method for doubly fed induction generator is proposed in this paper. Doubly fed induction generators are commonly used in wind turbine applications because of fast dynamic response and grid synchronizing capability. Dependency on speed measurement is an important drawback in wind turbine applications. A reduced order model based observer is used to estimate the torque, rotor flux and speed. The results of estimation are used in finite set model predictive control. The next step torque and rotor flux are predicted for all seven possible switching states. The best apply is that minimizes the cost function. The proposed method is qualified by simulation results.