Obtaining Partial Solutions for Divergent State Estimation Problems in Large Power Systems

This paper describes implementation of a multi-area state estimation approach where a converged solution can be obtained for a large part of the network even if the conventional state estimator does not converge. This is accomplished by locating and isolating the area which contains the gross error that is responsible for divergence. The paper investigates and addresses two main issues, namely area partitioning and network observability. Simulation results are provided to illustrate the effectiveness of the proposed approach.