Signal Frequency Estimation via Kalman Filter and Least Squares Approach for Non-uniform Signals

An innovative approach for estimating power system characteristics has been devised, and it is predicated on the Kalman filter (KF) and least squares (LS) approach. The procedure calculates an approximation of the signal’s frequency by taking three samples that are spaced equally apart. The computation of frequency requires the utilisation of the least squares method, which results in the identification of successive triples of instances. The difficulty can be solved by employing a slightly altered version of a method that calls for fewer mathematical calculations, but the resultant predictions are slightly less accurate. The strategy that has been suggested is the one that functions most effectively for predicting frequency. This is due to the fact that it has a fast response time and requires less data processing. In this work, we start by performing initial filtering using the KF before using the LS technique to estimate the frequencies. The responsiveness of the approach and its computation is precise. The discussion is on the factors that have an effect on the efficiency of such an approach, including the window size, sampling interval, distortion, harmonic components, and the filtering process. The performance results show that the suggested KF, along with the LS approach, produces a lesser estimate of the mean square error in comparison with other LS-based approaches.