Advantage of Unscented Kalman Filter over Extended Kalman Filter in dynamic state estimation of power system network

Abstract

The Kalman filter is a set of mathematical equations which are used to estimate the state of a system by minimizes the mean of the squared error. In this paper Kalman filtering is used for the estimation of states of IEEE 14 bus power system network. Here we considered dynamic states i.e. rotor angle in radians and speed in rad/sec of all the generators present in the system. We mainly focus on the advantages of Unscented Kalman Filter (UKF) over Extended Kalman Filter (EKF) by comparing both estimation methods. In EKF state distribution is approximated by a Gaussian Random Variable (GRV), which is then propagated analytically through a first order linearization of the non-linear system. This introduces errors in posteriori mean and covariance of the transformed GRV, which may lead to suboptimal performance of the filter. The UKF addresses this problem by using a deterministic sampling approach. In this paper these two algorithms are tested on an IEEE-14 bus,5-generator test system by applying test cases like sudden load change and configuration topology error to show how adaptive these filters during those conditions.