RECURSIVE STATE ESTIMATION ALGORITHM FOR ILL-CONDITIONED POWER-SYSTEMS

This paper describes a simple, reliable and very fast estimation algorithm for on-line analysis of ill-conditioned power systems. The estimator proposed is decoupled in nature and is suitable for on-line sequential processing of measurements telemetered, either one at a time or in batches. The conventional Weighted Least Square (WLS) method when solved iteratively by the normal equation method, exhibits a tendency to be numerically unstable on some networks. When the power system is ill-conditioned, it will manifest itself in the form of slow convergence or failure to converge. To overcome this difficulty the proposed algorithm is formulated as a Hessian approximation by Brown-Dennis technique of the information matrix in the WLS problem. As the non-zero elements of the Hessian matrix are not recomputed but are extracted from the Jacobian stored in computer memory, the additional computing time for implementing the new algorithm is quite small. Digital simulation results are presented for 13-bus and 43-bus ill-conditioned test power systems.