A Fourier-Based Phasor Estimator With a Modified Moving Average Filter and Its Application in Distribution Networks

Abstract

Phasor estimation is fundamental for most real-time analysis, monitoring, and control tasks in power systems. New applications on microgrids and active distribution networks make such estimation increasingly important, leading to many research efforts focusing on known approaches, such as the Fourier and cosine filters, to abate the estimation errors; others rely on newly applied algorithms like the Taylor–Kalman–Fourier filter (TKF). This variety has led to discussions about the application cases of each technique, normally demoting the Fourier filter (FF). In this spirit, the FF was reworked under the Hilbert space power theory, showing its conceptual correctness and then deriving an accurate and fast alternative relying on a discrete, lead-compensated filter. Such FF is tested over an active distribution network system, showing advantages in relevant scenarios, namely impedance estimation, grid-monitoring, and fault location. The proposal exhibits better dynamic performance and overshoot than the conventional techniques, with no increased complexity, and “cleaner” results than the TKF, which is, however, faster to estimate the system's impedances.