With the improvement of the degree of interaction of modern power systems, it is of great significance for the safe operation of the system to fully and accurately grasp the real-time operating status of the system. The power system state estimation can ensure the validity and accuracy of the data obtained by the power grid. Dynamic state estimation has been widely used because it can predict the operating trend of the system. Dynamic state estimation is mainly based on the principle of Kalman filtering, and predicts the state quantities of different time sections by measuring and updating. However, Kalman-type filters are susceptible to unknown noise, resulting in poor filtering performance. Aiming at the strong non-Gaussian character of unknown measurement noise in power system, low estimation accuracy, poor tracking performance and weak robustness of power system state estimation, an improved strong tracking unscented Kalman filter (MSTUKF) is proposed. The strong tracking filter makes the unknown measurement noise approximately satisfy the Gaussian distribution by forcing the residual sequence to be orthogonal, At the same time, a multiple fading factor is introduced into the prediction error covariance to reduce the weight of the measurement noise，thereby reducing the influence of the unknown measurement noise on the unscented Kalman filter. The filtering efficiency can be improved by correcting the quantity measurement and the state quantity in real time through the outlier elimination algorithm based on the sliding window before filtering. The simulation results show that compared with the traditional UKF method, the MSTUKF proposed in this paper improves the accuracy of state estimation, and has better filtering effect and higher robustness than UKF in the presence of unknown measurement noise in the system.
 

state estimation; unscented Kalman filter; strong tracking; multiple fading factors