Purpose
The purpose of this paper is to propose an extended Kalman particle filter (EPF) approach for dynamic state estimation of synchronous machine using the phasor measurement unit’s measurements.
Design/methodology/approach
EPF combines the extended Kalman filter (EKF) with the particle filter (PF) to accurately estimate the dynamic states of synchronous machine. EKF is used to make particles of PF transfer to the likelihood distribution from the previous distribution. Therefore, the sample impoverishment in the implementation of PF is able to be avoided.
Findings
The proposed method is capable of estimating the dynamic states of synchronous machine with high accuracy. The real-time capability of this method is also acceptable.
Practical implications
The effectiveness of the proposed approach is tested on IEEE 30-bus system.
Originality/value
Introducing EKF into PF, EPF is proposed to estimate the dynamic states of synchronous machine. The accuracy of a dynamic state estimation is increased.
Observability and detectability analyses for dynamic state estimation of the marginally observable model of a synchronous machine
