State Estimation Using the Unscented Kalman Filter in Nodal Redundancy for EPS

The integration of renewable energy sources, decentralized generation, and increasing grid complexity have motivated the development of advanced state estimation strategies for power systems. This paper presents a distributed state estimation (DSE) framework based on the Unscented Kalman Filter (UKF) and nodal redundancy partitioning. Unlike traditional centralized estimators, the proposed method divides the electrical network into interconnected clusters, each executing a local UKF and exchanging information across boundary nodes to reconstruct the global state. The methodology was validated on the IEEE 14-bus test system and compared against centralized and distributed implementations of the WLS, EKF, and UKF estimators. Performance was evaluated in terms of estimation accuracy, convergence, and computational efficiency. Results show that the distributed UKF achieves improved accuracy in nonlinear scenarios and reduces computational time by up to 30% compared to centralized implementations. This study demonstrates the feasibility and benefits of integrating UKF with nodal redundancy for real-time, scalable state estimation in modern power systems.