Dynamic state estimation of power system using hybrid whale-tunicate optimized unscented Kalman filter based on wide area measurement systems

The Energy Management Systems (EMS) utilizes Dynamic State Estimation to monitor the operating condition of the power grid. For a synchronous generator, the speed and angle of a rotor are the most important dynamic states. These states are responsible for producing uncertainty in the power system models. Extended Kalman Filter (EKF) and Unscented Kalman Filters (UKF) are utilized to estimate the dynamic states of the power system when it perceives significant disturbances or noise. The conventional Unscented Kalman Filter (UKF) has limitations in terms of slow convergence speed and its lack of positive semi-definiteness. This prohibits its ability to accurately track the transients of the system. The scaling parameters of UKF (alpha and beta) are optimally tuned to overcome these issues. In this paper, a novel Hybrid Whale-Tunicate-based UKF (HW-TO-UKF) is proposed for tuning the scaling parameters of UKF for reducing the mean difference between the real state and the measured state. During the validation of the proposed method, better robustness and statistical efficiency are achieved by defining the objective function as the minimization of the error difference between the actual and measured states. The simulation is carried out on the WSCC-3 and NPCC-48 machine systems to demonstrate the performance of the proposed HW-TO-UKF over the existing models.