Observability analysis and distributed state estimation

This paper presents a power system state estimation (SE) in a distributed manner. In order to perform a state estimation algorithm we must ensure system's observability by restoring local and global observability. In the proposed method, measurements are divided into local and boundary. Each control area performs its own local state estimation using local measurements and exchanges information, concerning boundary states and measurements, with a coordination SE. The coordinator computes the overall system state vector and decides whether the SE algorithm converges or not. When every single area converges, the algorithm terminates and the system wide state vector is produced. Observability analysis and distributed state estimation algorithms are tested with the IEEE 14-bus system, and two realistic systems in Rhodes Island. Results concerning Rhodes's system are presented in section 4.