Power System Fundamental Frequency and Harmonic Estimation using Improved LMS

Abstract

Maintaining good power quality is an important challenge in modern power systems, as disturbances can adversely affect the performance and reliability of electrical equipment and may lead to significant economic losses. Among the commonly occurring power quality disturbances, harmonics, voltage sag, and voltage swell are particularly important for applications such as power quality monitoring, system protection, economic dispatch, and security assessment. Accurate estimation of the fundamental frequency and harmonic components is therefore essential for reliable power system operation. Several techniques have been developed for this purpose. The Fast Fourier Transform (FFT) is one of the conventional approaches; however, its performance can be affected by limitations such as spectral leakage and the picketfence effect. The Least Mean Square (LMS) algorithm provides a relatively simple computational approach for estimating the fundamental frequency and harmonic components and can be implemented with comparatively low computational complexity. In this study, the performance of the LMS algorithm is investigated for fundamental-frequency estimation under different signal-to-noise ratio (SNR) conditions. The analysis focuses on its ability to provide reliable frequency estimation in the presence of varying levels of noise, which is important for practical power quality monitoring applications.