Neural Network-Based Enhanced PI and Higher-Order Sliding-Mode Control for Harmonics Mitigation

Abstract

Due to the varying load conditions in electrical networks, typical sliding mode control techniques are insufficient to control active power filters (APFs). Consequently, this study proposed a higher-order sliding mode and an enhanced PI technique to enhance the performance of the APF. This controller has been chosen because of its ability to respond to network parameter variations and interference of the network parameters. The results reveal that this approach gives outstanding results, with a rapid voltage stabilization time of 0.06 seconds after > 0.4 seconds. Furthermore, total harmonic distortion (THD) has been greatly reduced to 1.89%, meeting the IEEE 519 standard for reliability in electrical distribution networks.