Computation of power quality symmetrical components using fuzzy logic based linear combiners

This paper presents a new approach to estimate the symmetrical and α-β components of the voltage and current signals of a power distribution network using Fourier linear combiners. The learning parameter for updating the weights of the combiner is adapted using a fuzzy logic-based algorithm obtained from the error covariances. The new algorithm is tested using recorded data from an industrial load bus supplying power to electronically controlled induction furnace loads. The estimation results are compared with a 19^th order Kalman filter approach to highlight the accuracy and fast tracking capabilities of the new approach. The linear combiner, however, presents a simple and implementable algorithm in comparison to the well known Kalman filter approach