ADVANCED CHARACTERIZATION OF POWER SYSTEM HARMOICS USING MUSIC ALGORITHM ON NON-STATONARY WAVEFORMS

Maintaining power quality in modern power systems is crucial, as the integration of renewable energy sources and the rise of non-linear loads have resulted in intricate harmonic distortions. Ordinary methods based on Fourier series, like the Fast Fourier Transform (FFT), are frequently used for harmonic analysis; however, they often struggle to accurately analyze non-stationary signals, which are commonly observed due to system disturbances and variations. This research investigates the Multiple Signal Classification (MUSIC) algorithm as a more advanced option for harmonic analysis in power systems. The high-resolution frequency estimation provided by MUSIC allows for precise identification of harmonic amplitudes and phases, even in the presence of noise and fluctuations in the fundamental frequency. In this study, synthetic non-stationary waveforms that simulate actual conditions at the Point of Common Coupling (PCC) are examined to demonstrate MUSIC's capability to identify harmonics accurately where traditional FFT-based methods fail. Findings from simulations indicate that MUSIC not only provides enhanced accuracy in harmonic characterization but also exhibits resilience under non-stationary conditions, making it particularly effective for real-time power quality monitoring. This research emphasizes MUSIC as an important progress in harmonic analysis, presenting a valuable resource for enhancing power system reliability in dynamic settings.