EKF-and wavelet-based algorithms applied to harmonic detection for active shunt filters

Abstract

This work presents a performance comparison among algorithms able to estimate in real-time the fundamental component of the current absorbed from the supply by a nonlinear load. This fundamental component can be used to calculate the current reference for an active shunt filter (ASF). The effectiveness of ASFs in reducing the harmonic contents of the supply currents is strictly related to the ability of an algorithm to give a reliable and accurate estimate of the harmonic contents of the load current both at steady-state and in transient conditions. We have considered algorithms based on the use of a synchronous rotating frame (SRF), wavelet analysis, and Kalman filter theory. In order to overcome the problem related to the choice of the covariance matrices of the Kalman filter-based algorithm we propose to use a genetic algorithm (GA) to perform an off-line automatic search for optimal parameters. Experimental results are presented considering a three-phase induction motor drive as nonlinear load.