This study introduces a novel fractional-order sliding mode control (FOSMC) strategy for DFIG-based wind turbines, aiming to optimize energy conversion efficiency. The approach combines FOSMC with radial basis function neural networks (RBFNNs) to achieve a fast practical finite-time convergence and eliminate chattering through a specially designed nonsingular sliding surface. System robustness is reinforced by a RBFNN-driven real-time uncertainty estimation. Simulation results confirm the controller’s high tracking precision, strong robustness against uncertainties, and complete suppression of chattering effects.
Boudjemia et al. (Wed,) studied this question.