Distributed Dynamic State Estimation of Power Systems

In this paper, a novel distributed dynamic state estimation (DSE) method for real-time monitoring of power systems is implemented. In modern large-scale power grids, the number of deployed meters and the frequency of collecting data have remarkably increased. Such a growth in the spatiotemporal size of collected data overwhelms the existing monitoring system with a centralized star structure. Streaming of data from all meters of the network to the central control center also increases communication latency. To overcome these challenges, the power system is partitioned into subsystems with local estimators, and the DSE process is distributed among local estimators. The distributed DSE is hosted at regional control centers and utilizes distributed extended Kalman filtering based on internodal transformation theory to estimate the dynamic states of power systems. The local estimators only require data within their own subsystem, and information is exchanged only between neighboring subsystems. The proposed distributed DSE is implemented on a 68-bus test system, and its accuracy is demonstrated.