ADAPTIVE LINEAR ALGORITHMS FOR POWER SYSTEM FUNDAMENTAL AND HARMONIC ESTIMATION

Frequency and harmonics are two essential parameters for power system functions
such as power quality monitoring, power system protection, economic dispatch and
security assessment. Several approaches have been presented in the last two decades
to estimate the fundamental frequency and harmonics. The most popular algorithm
that has been implemented to estimate, and quantify power system fundamental
frequency and harmonic components is the Fast Fourier Transform (FFT). However,
this technique has a few negative implications such as spectral leakage and picketfence
effect. On the other hand, Least Mean Square (LMS) algorithm is known for its
ease of structure, computation and simplicity. Hence, one of the objectives of this
thesis is to address and enhance the introduced fundamental frequency adaptive filter
method which was based on modified variable step size LMS (MVSS) algorithm
using generalized square error normalized LMS algorithm. This approach is aimed at
reducing the number of parameters and steps in MVSS as well as to improve the
convergence rate, quick tracking and low Mean Square Error (MSE). Moreover, in
order to attain faster response and more harmonic estimation accuracy, a second order
Recursive Levenberg Marquardt (RLM) harmonic estimator is presented.