Abstract
The unscented Kalman filter (UKF) has been widely used in power system dynamic state estimation (DSE), which provides a guarantee for the establishment of a high-quality database to store power network information. However, traditional UKF faces two main problems. First, the performance of traditional UKF will deteriorate due to the interference of non-Gaussian noise. Second, the traditional UKF estimator will be interrupted owing to the Cholesky decomposition of asymmetric positive definite matrix. To deal with these problems, this paper develops a square root UKF based on minimum error entropy with fiducial points (MEEF) criterion (SR-MEEF-UKF). The MEEF criterion inherits the common advantages of correntropy and error entropy, exhibiting robustness to outliers. At the same time, the kernel matrix in SR-MEEF-UKF is non-singular. Through performing DSE in the IEEE-14 bus system, the proposed SR-MEEF-UKF not only has excellent performance in non-Gaussian noise environments, but also has strong numerical stability.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call