Research on Improved Kalman Filtering and Robust Identification for Dynamic Power System State Estimation

This paper presents a distributed dynamic state estimation approach specifically designed for large power systems. To enhance scalability and efficiency, the system is divided into several overlapping regions, where each region independently performs state estimation. The proposed method utilizes information fusion to maintain consistency in the state estimates, particularly at boundary nodes shared between regions. In addition, acknowledging the potential impact of non-Gaussian noise and abnormal measurements on power system monitoring devices, the paper integrates M-estimation theory with unscented Kalman filter to improve the robustness of the estimation process. This integration strengthens the estimator's capacity to mitigate measurement anomalies, ensuring more reliable and accurate state estimates across the power grid. This methodology shows promise in creating a more resilient and scalable state estimation framework for modern, complex power systems, addressing key challenges related to measurement errors and computational efficiency.