Parameter Estimation Techniques Applied to Power Networks

Abstract

This paper represents the estimation of frequency which is an important power system parameter by several variant of recursive techniques. Estimation of frequency is not new as found in the literature such as Discrete Fourier Transformation, Least Mean Square Technique etc. In this paper we have implemented Recursive least square (RLS) and Extended Least square (ELS) for estimation of frequency. These are well known algorithms for their simplicity in computations and good convergence properties. Such algorithms require three step calculations with estimation of amplitude and phase followed by frequency. They are simple and attractive with the implementation of covariance matrix. Again the choice of covariance matrix is very crucial without which there might be delay in convergence for estimation of power system parameters. At the same time, feasibility of the above algorithms is tested with signal buried with noise. The above work can be extended for real time implementation, which will be immensely helpful for Modern Power System engineers.