Position Sensorless Predictive Power Control of Doubly-fed Induction Generator Based on Rotor Current Model Reference Adaptive System

Doubly-fed induction generators have been widely used in wind power generation systems due to their low converter power, flexible active and reactive power control, and two way flow of energy. The existing vector control needs to tune multiple PI coefficients, and additionally needs to install a position sensor to obtain the rotor position and speed, which increases the volume and cost of the generator while reducing the reliability of the system. Therefore, a new solution that does not require PI setting and installation of position sensors is proposed. First, on the basis of analytically deriving the complex power differential on the stator side, the voltage reference value on the rotor side is directly calculated based on the idea of power deadbeat control without any controller setting work. Secondly, by comparing the measured rotor current with the estimated rotor current based on the stator flux current model, a position estimation method based on the rotor current model reference adaptive is constructed, thereby realizing the position sensorless control. The experimental results show that the method proposed can track the actual position and speed well under various working conditions such as sub-synchronization, super-synchronization and power step, and has excellent steady-state performance and fast dynamic response and provides the doubly-fed wind power generation system provides a low-cost, high-performance, and easy-to-apply solution.