A genetic based algorithm for frequency-relaying applications

In this paper, a new method for on-line tracking of power system frequency is presented. The method is based on genetic algorithms (GAs) optimization technique. GAs are optimization and learning techniques based on natural selection and natural population genetics. The problem is formulated as an estimation problem. GAs with different fitness functions, that evaluate the solution are presented. Simulated examples are used to test the proposed method. Different voltage waveforms with different frequency deviation from the nominal value as well as different magnitudes are generated to perform the study. The bus voltage waveform in each case is sampled at a pre-selected rate to form the digital data set. The effects of sampling rate, sample window size as well as the effects of GAs parameters and operators, such as population size, crossover and mutation probabilities are studied. Results are reported and discussed. (C) 2000 Elsevier Science S.A. All rights reserved.