Robust Dynamic State Estimation for Power System Based on Generalized Correntropy Loss Function and Unscented Kalman Filter
Tengpeng Chen,Hongxuan Luo,Yuhao Sun,Eddy Y. S. Foo,Tao Zeng,Gehan A. J. Amaratunga
DOI: https://doi.org/10.1002/tee.24229
IF: 0.923
2024-11-27
IEEJ Transactions on Electrical and Electronic Engineering
Abstract:Dynamic state estimation (DSE) of the power system is one of the most important means to ensure a safe and stable operation of the power system. However, the electromagnetic field environment, communication noise and equipment failure often lead to the appearance of non‐Gaussian noise and bad data, which ultimately cause performance degradation of traditional DSE algorithms based on the mean‐square error criterion. In this paper, a DSE method based on the generalized correntropy loss (GCL) function is proposed, which is well adapted to estimate the dynamic characteristics of the power system. The new robust DSE method can effectively deal with problems such as non‐Gaussian noise and bad data, thus improving the state estimation accuracy. Simulation results carried out on the IEEE 39‐bus system with synchronous generators demonstrate the robustness of the proposed DSE method. © 2024 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.
engineering, electrical & electronic