Power System Fusion State Estimation Based on a Switched System Model: Comparison Between EKF and UKF

This paper makes a comparison study between the extended Kalman filter (EKF) and the unscented Kalman filter (UKF) on the fusion state estimation of power systems. The power system SE is based on a switched system model which comprises two subsystems. One subsystem has a simple form with a identity state transition matrix and zero input. The other is derived from the linearization of the power flow equation. A switching rule based on innovations is established through utilising the phasor measurement unit with high precision. Theoretically, the UKF has one or two more order accuracy than the EKF which commonly uses first-order linearization. Nevertheless, the simulation on the IEEE test system demonstrates that the process noise covariance matrix has a significant influence on the estimation performance of UKF and EKF both.