Signal-dependent preprocessing of buffered PMU measurements for hybrid state estimation

This paper presents a signal-dependent scheme for selecting the length of the averaging window for PMU measurements to improve the performance of the power system state estimator. The method is based on the Shewhart method for detecting mean shifts in the buffered data. We suggest ways to tune the parameters of the scheme by deriving expressions for the probability of false alarm and the probability of using the entire buffer given all buffered data have the same mean. We also propose a method to handle oscillatory transients captured by the buffered PMU measurements, which may otherwise degrade the performance of the buffering scheme. Simulations performed on the IEEE 14-bus test system show that the proposed method outperforms an existing state-of-the-art method.