Dynamic state estimation based synchronous generator model calibration using PMU data

Abstract

A method for synchronous generator parameter estimation by dynamic state estimation is presented using PMU data. A three phase physically based two-axis generator model with two damper windings is utilized for the dynamic state estimation. Eleven independent parameters are estimated: ten self and mutual inductances of the various windings and the inertia constant. These parameters can be either used for the three phase physically based model or be converted to the usual dq0 synchronous generator model. The proposed method requires time domain three phase sampled data for better estimation accuracy. In the paper, the PMU phasor data with disturbance provided by NERC was converted to three phase sampled data first and then they were fed into the dynamic state estimation. The comparison between the calibrated parameters and the actual parameters provided by NERC is shown to prove the effectiveness of the proposed method. The accuracy of the proposed method can be greatly improved if the three phase sampled value data are provided at the generator three phase terminals, the exciter and the governor.