Frequency-Domain System Identification of a First Order Governor-Turbine Model from PMU Ambient Data

A first-order or second-order model of governor-turbine (GT) can be utilized in dynamic state estimation (DSE) of synchronous generators (SGs). The input and output of the GT system are rotor speed deviation and mechanical power of SGs. An accurate GT model is essential to achieve a reliable DSE. In this paper, an identification method of a first-order GT model is proposed in the frequency domain by using PMU ambient data. This method exploits low-frequency dynamics of the GT system, which enables the mechanical power to be replaced with electrical active power of SGs in a low-frequency range. The method is designed to be performed on average periodograms of the active power and rotor speed deviation, which are calculated from SG voltage and current phasors. The average periodogram can overcome a large variance of periodogram and noise effects in PMU data. The usefulness of the proposed method is demonstrated through simulations of the IEEE 39-bus system.