A kalman filter‐based technique for combined digital estimation of voltage flicker and phasor in power distribution systems

Electric power quality is of prime importance when sensitive electronic equipment is involved. One of the aspects of power quality is known as flicker level. The present paper proposes an algorithm for real-time monitoring of voltage flicker based on the Kalman filtering technique. First a stochastic state space model of the system voltage with a linear time-invariant transition matrix has been established, yet non-linear in the scalar observation model. Using noisy measurements of a single-phase voltage, an extended Kalman filter leads to instantaneous value of rms voltage and phase. Then the flicker level has been derived by appropriate filtering and statistical processing on short records of the rms voltage measurements. Simulated and actual data demonstrating the effectiveness of the overall scheme in estimating voltage phasor and flicker for a wide range of swing frequencies are presented and discussed. Among the several power system applications of the new stochastic observer considered, a remote load control concept based on the carrier amplitude modulation is shown being possible and included are test results confirming the filter capability to retrieve sequences of amplitude-modulated bit patterns travelling through the power line to customer sites.