Research On Sensorless Control Strategy Of DFIG Wind Energy Conversion System

Nowadays, energy shortage and environmental pollution are the important factors that restrict the development of economy. It’s urgent to develop the utilization of renewable energy. Because of its large reserves and wide distribution, wind energy is very important among these renewable energies. In the existing wind power generation systems,the Doubly Fed Induction Generator(DFIG) system is widely used because of its small size and low cost. For improving the wind energies’ coefficient of utilization and the generating efficiency of wind turbine generators, papers discuss the maximum wind energy tracking control of DFIG system, actually this method controls the DFIG’s power which is the derivate of the wind and rotor speed. Rotor speed detection, which is used to achieve with sensor, is very important for the DFIG’s control system. But with the use of sensor, the cost of the system may increase, and reliability decrease. Therefore, on the basis of DFIG’s maximum wind energy tracking control method, this paper deeply studies the sensorless control ways of the system.The sensorless control strategies are applicated on the Rotor-Side Converter(RSC) of the DFIG system, and also set up the experimental platform. The operational principle and math model of DFIG and RSC in the system are analysised in this part. On this basis, this paper uses the Stator Voltage Oriented(SVO) control strategy setting up the double loop control system of power and current, to achieve the maximum wind energy capture and the decoupling control of power. When the wind changes, the DFIG system’s simulation model can also realize the real time tracking of wind energy.Using the speed sensor may let the concentric angle to be uncertaint, the reliability to be low, so the paper discusses the estimated speed method of Model Reference Adaptive System(MRAS) deeply. The Rotor-Current-Based MRAS Observer(RCMO) can estimate the actual rotor speed, but has some error. So the paper studies Torque-Norm-Based MRAS Observer(TNMO) in detail. When speed identificates, using the low pass filter replace the pure integral, also compensating the phase and amplitude. The second method for estimating rotor speed has higher accuracy through theoretical analysis and simulationverification. Using this estimated speed for the sensorless control is more successful.Finally, MATLAB/Simulink simulation of the sensorless control strategies in the DFIG system is correctly performed. After this, experiments of the TNMO speed estimated method are confirmed on the d SPACE platform. Experimental results verifiy the feasibility and correctness of the adopted strategies.