Synchrophasor Estimation for Three-Phase Systems Based on Taylor Extended Kalman Filtering

Synchronized phasor and frequency measurements are key tools for the monitoring and management of modern power systems. Under dynamic conditions, it is vital to define algorithms that allow accurately measuring time-varying signals with short latencies and high reporting rates. A dynamic phasor model can help the design of these algorithms and, in particular, of those based on the Kalman filter approach. This article proposes a three-phase synchrophasor estimator based on the extended Kalman filter; state variables are obtained from Taylor expansions of amplitudes and phase angles. The underlying dynamic model considers the inherent relationship among the phases and includes harmonics in an effective way. The process noise covariance matrix that allows representing the uncertainty introduced by the dynamic model has been written by considering that practical ac power systems are nearly three-phase symmetric during typical operation. This a priori information allows improving noise rejection and increasing accuracy in the presence of amplitude modulation, as highlighted by the reported simulation results.