A combined strategy of torque error feed-forward control and blade-pitch angle servo control is proposed to improve the dynamic power capture for wind turbine maximum power point tracking (MPPT). Aerodynamic torque is estimated using the unscented Kalman filter (UKF). Wind speed and tip speed ratio (TSR) are estimated using the Newton–Raphson method. The error between the estimated aerodynamic torque and the steady optimal torque is used as the feed-forward signal to control the generator torque. The gain parameters in the feed-forward path are nonlinearly regulated by the estimated generator speed. The estimated TSR is used as the reference signal for the optimal blade-pitch angle regulation at non-optimal TSR working points, which can improve the wind power capture for a wider non-optimal TSR range. The Fatigue, Aerodynamics, Structures, and Turbulence (FAST) code is used to simulate the aerodynamics and mechanical aspects of wind turbines while MATLAB/SIMULINK is used to simulate the doubly-fed induction generator (DFIG) system. The example of a 5 MW wind turbine model reveals that the new method is able to improve the dynamic response of wind turbine MPPT and wind power capture.

In order to improve the prediction accuracy of wind speed, this paper proposes a hybrid wind speed prediction (WSP) method considering the fluctuation, randomness and nonlinear of wind, which can be applied to short-term deterministic and interval prediction. Variational mode decomposition (VMD) decomposes wind speed time series into nonlinear series Intrinsic mode function 1 (IMF1), stationary time series IMF2 and error sreies (ER). Principal component analysis-Radial basis function (PCA-RBF) model is used to model the nonlinear series IMF1, where PCA is applied to reduce the redundant information. Long short-term memory (LSTM) is used to establish a stationary time series model for IMF2, which can better describe the fluctuation trend of wind speed; mixture Gaussian process regression (MGPR) is used to predict ER to obtain deterministic and interval prediction results simultaneously. Finally, above methods are reconstructed to form VMD-PRBF-LSTM-MGPR which is the abbreviation of hybrid model to obtain the final results of WSP, which can better reflect the volatility of wind speed. Nine comparison models are built to verify the availability of the hybrid model. The mean absolute percentage error (MAE) and mean square error (MSE) of deterministic WSP of the proposed model are only 0.0713 and 0.3158 respectively, which are significantly smaller than the prediction results of comparison models. In addition, confidence intervals (CIs) and prediction interval (PIs) are compared in this paper. The experimental results show that both of them can quantify and represent forecast uncertainty and the PIs is wider than the corresponding CIs.