Local and Wide-Area PMU-Based Decentralized Dynamic State Estimation in Multi-Machine Power Systems

TLDR: In this paper, the authors employ the extended Kalman filter with unknown inputs, referred to as the EKF-UI technique, for decentralized dynamic state estimation of a synchronous machine states using terminal active and reactive powers, voltage phasor and frequency measurements.

Abstract: Accurate measurement of the rotor angle and speed of synchronous generators is instrumental in developing powerful local or wide-area control and monitoring systems to enhance power grid stability and reliability. Exogenous input signals such as field voltage and mechanical torque are critical information in this context, but obtaining them raises significant logistical challenges, which in turn complicates the estimation of the generator dynamic states from easily available terminal phasor measurement unit (PMU) signals only. To overcome these issues, the authors of this paper employ the extended Kalman filter with unknown inputs, referred to as the EKF-UI technique, for decentralized dynamic state estimation of a synchronous machine states using terminal active and reactive powers, voltage phasor and frequency measurements. The formulation is fully decentralized without single-machine infinite bus (SMIB) or excitation model assumption so that only local information is required. It is demonstrated that using the decentralized EKF-UI scheme, synchronous machine states can be estimated accurately enough to enable wide-area power system stabilizers (WA-PSS) and system integrity protection schemes (SIPS). Simulation results on New-England test system, Hydro-Quebec simplified system, and Kundur network highlight the efficiency of the proposed method under fault conditions with electromagnetic transients and full-order generator models in realistic multi-machine setups.