Recursive hierarchical estimation of the states of an electrical power system

Abstract A three-level, hierarchical state estimation method for electrical power systems is proposed, which allows the states of both the network and those of the generating units to be sequentially estimated. The states of the network level are estimated by a two-level, linear, hierarchical, optimal filter, whereas those of the lower generating level by the usual extended, discrete-time Kalman filters. The proposed decentralized method is based on a suitable decomposition of the overall system. Besides having the advantage of using several coordinated estimators of low dimensionality, the described method also provides for a simple network level disturbance detection and localisation procedure. A numerical example illustrates the proposed state estimator.