Discrete-time decentralized estimators for interconnected systems

The authors propose a method for designing discrete-time decentralized state estimators for interconnected large-scale systems. The interface variables, which can be considered as a stochastic input to the subsystems, are measured with a least-squares estimator. This stochastic input is replaced by the estimate and a noise term. Local Kalman filters are then designed using local measurements to estimate local states. A two-area power system example is shown to illustrate the effectiveness of the proposed method.<>