Estimation of power system frequency

Abstract

Frequency is an important parameter in power system for indicating the dynamic balance between amount of power generated and amount of power consumed. Small changes in system frequency can affect the operation of a power system in a great manner. Thus, estimating the frequency is one of the most important task in the power system operation. A least mean square (LMS) algorithm with varying step size to estimate the frequency is studied in this report and its performance under the influence of harmonics, noise, unbalanced conditions, step changes in frequency, etc is observed through various simulations. Next, Frequency estimation is carried out with an improved Non-linear Least Squares (NLS) method in which frequency is estimated through a 1-D search over a specified range. The validity of method under different conditions including line to ground fault is also tested. This method is also verified using the data obtained from an experimental setup consisting of a non-linear load.