Aiming at the problem that the doubly-fed wind turbine system is susceptible to grid voltage drops, resulting in speed estimation accuracy and system instability of traditional MRAS,  a speed estimation strategy for Doubly Fed Induction Generator (DFIG) based on sliding mode adaptive control under power grid failure is   proposed in this paper. This method improves the controllability of DFIG under grid faults through a self-demagnetization strategy, and improves the current tracking and flux observation accuracy. The stator flux voltage model of DFIG is used as a reference model, the stator flux current model is used as an adaptive model. The Second-Order Generalized Integrator based Quadrature Signal Generator (SOGI-QSG) is used to achieve positive and negative sequence separation of the stator flux. A gain normalized FLL is used to make the output signal of the SOGI-QSG unaffected by the stator flux amplitude. The harmonics of the stator flux are eliminated by cascading low-pass filters, and the stator flux positive sequence component is used as the reference value of the estimator, which solves the problem of inaccurate DFIG speed estimation by traditional Sliding Mode MRAS (SM-MRAS) when the stator voltage drops. At the same time, the sliding mode gain coefficient is adjusted in real time according to the stator flux positive sequence error, which improves the accuracy of speed estimation during high-speed operation and enhances system stability. Finally, the performance of the proposed control method is verified based on Matlab/SimuLink.
 

Doubly Fed Induction Generator ;,speed estimation,grid voltage dip,model reference adaptive