Dynamic State Estimation for Power System with Communication Constraint Using Event-Triggered Cubature Kalman Filter

Accurate dynamic state estimation plays an important role in power systems. Although various filtering algorithms, such as unscented Kalman filter (UKF) and particle filter (PF), have been proposed based on phasor measurement units (PMUs), they occupy a huge communication bandwidth without specific concern, which puts heavy burden on the communication network especially when wireless communications are widely applied in smart grids. In order to relieve this communication burden, the event-triggered cubature Kalman filter (ETCKF) is proposed based on the stochastic event-triggered schedule in this paper, which guarantees that only the measurements containing innovational information are transmitted and has advantages over other event-triggered schedules, such as the deterministic event-triggered schedule, since it can maintain the Gaussian property of the conditional distribution of the system state. Based on the developed nonlinear event-triggered schedule, the cubature Kalman filter (CKF), using the third-degree spherical-radial cubature rule, further provides more accurate estimation than UKF and has lower computational complexity than PF. The proposed filter can effectively reduce the communication rate while ensuring the accuracy of filtering. Finally, the standard IEEE 39-bus system is utilized to verify the feasibility and performance of the proposed method.