Intelligent Adaptive Super-Twisting Sliding Mode Control With RBF-Gain and ANN Sign Approximation for DFIG-Based Wind Energy Systems

Abstract

This paper presents a novel intelligent control strategy for the rotor-side converter (RSC) of a doubly-fed induction generator (DFIG) in wind energy conversion systems (WECS), integrating radial basis function (RBF) neural network-based adaptive gain tuning with an artificial neural network (ANN)-based sign-function approximation within the framework of a super-twisting sliding mode controller (STSMC). The proposed RBF–ANN STSMC dynamically adapts control gains in real-time based on the system tracking errors and smoothly approximates the discontinuous signum function to suppress chattering without compromising robustness. A Lyapunov-based stability analysis ensures finite-time convergence under system uncertainties and external disturbances. Extensive simulations in MATLAB/Simulink under diverse scenarios, including wind-speed transients, voltage dips, load disturbances, and parameter variations, demonstrate the superiority of the proposed controller over the standard STSMC controllers. Performance evaluations based on tracking accuracy, settling time, total harmonic distortion (THD), and smoothness of the control effort confirm enhanced robustness, disturbance rejection, and compliance with power quality standards. The integrated RBF–ANN architecture offers a promising and scalable solution for next-generation intelligent wind energy control systems.