Reduced-order estimation of power system harmonics

The estimation problem of unknown and time-varying power system harmonic signal amplitudes is addressed using a modified form of a signal model which has previously appeared in the literature. Using state-variable techniques, this signal model not only accounts for the unknown variations in signal amplitude but also for the uncertainty or measurement error introduced by variations in instrument transformer frequency response and burdens, and other non-ideal effects which are in the guise of measurement noise. A novel reduced-order state estimator is then used to identify the harmonic components of each measured signal for possible use in harmonic mitigation methodologies. Simulation results using this new harmonic signal model and associated reduced-order estimator are compared to those obtained by the Kalman filter. The main application of this contribution is to flexible AC transmission systems (FACTS) controllers (e.g., static condenser (STATCON) or dynamic voltage restorer (DVR)) which are capable of real-time reactive power and harmonic compensation.