Neural Network Complementary Sliding Mode Current Control of Active Power Filter

This study proposed an adaptive Radical Basis Function (RBF) neural control strategy with a complementary sliding mode approach to compensate the harmonic current in an Active Power Filter (APF). A backstepping algorithm is incorporated to simplify the design procedure. Meanwhile, a complementary sliding surface is employed to replace the standard sliding surface to eliminate the chattering. A neural estimator is designed to approximate the upperbound of the lumped nonlinearities in the APF. A simulation and real-time prototype using TMS320F28335 was built to demonstrate the validity of the proposed controller.