Recurrent Neural Network Global Sliding Mode Control Based on Fractional-order for Active Power Filter

In this study, a recurrent neural network global sliding mode control scheme based on fractional-order (RNNFGSMC) is introduced to achieve better performance for a three-phase active power filter (APF). The proposed RNNFGSMC scheme combines a fractional-order control method, a global sliding mode control method and a recurrent neural network structure. The primary objective of the proposed RNNFGSMC scheme is to construct a fractional-order global sliding mode controller. The RNN estimator is employed to approximate the unknown nonlinear function of the APF online. Simulation results are presented to show the effectiveness of the proposed scheme.