SUPPRESSION OF HARMONIC DISTORTIONS IN ELECTRIC POWER NETWORKS USING ARTIFICIAL INTELLIGENCE

Abstract. Introduction. This article proposes control methods based on artificial intelligence (AI) to reduce harmonic distortions in power supply systems. In modern energy systems, the increase in nonlinear loads and inverters leads to a rise in harmonics. Traditional passive and active filters are not sufficiently effective under varying load conditions. Therefore, the use of artificial intelligence tools for detecting and suppressing harmonics is considered highly relevant. Materials and Methods. Possibilities for real-time detection, prediction, and suppression of harmonic signals were developed using artificial neural networks (ANN), fuzzy logic control (FLC), and reinforcement learning (RL) algorithms. The research was carried out in the MATLAB/Simulink software environment. The models considered load variability, voltage waveform distortion, and control of harmonic components. The efficiency of the control algorithms was evaluated based on the Total Harmonic Distortion (THD) index. Results. Simulation results showed that control systems based on artificial intelligence, especially FLC and ANN models, are more efficient than traditional passive filters. The THD level was reduced by up to 30–40%. Adaptive control systems based on RL demonstrated their ability to respond to rapid changes in load. Conclusion. According to the research results, the application of artificial intelligence-based algorithms in reducing harmonic distortions in power supply systems yields high performance. Such systems are characterized by real-time operation, adaptability, and predictive capability. This is of great importance for ensuring the stability of modern power grids.