Frequency Estimation of Unbalanced Three-Phase Power Systems Using the Modified Adaptive Filtering

Abstract

In this paper, the problem of frequency estimation using adaptive filters is addressed based on the augmented complex normalized least mean squares (ACNLMS) technique. In other words, motivated from ACLMS technique, a new version called ACNLMS that outperforms the previous techniques is proposed. The proposed method makes use of two weight coefficients, like ACLMS, together with normalized time variant step size, thus its promising in increasing and enhancing both convergence rate and accuracy. Then, using derived weight coefficient, the frequency is estimated at each step. The performance and convergence analysis of the proposed method along with simulation results comparing with two existing techniques, CLMS and ACLMS, are provided in critical cases such as various unbalanced conditions and presence of harmonic distortion. Simulation results indicate that the proposed technique achieves a better performance in terms of convergence rate and frequency estimation accuracy as compared with CLMS and ACLMS techniques. Moreover, ACLMS achieves a smaller error variance than the other mentioned cases.