Fault-tolerant decentralized PCT estimators applied to non-linear electrical power system state estimation

The reliability of the power system state estimation depends on the availability of the measurements, which may be affected by hardware faults, especially sensor failures. In this work the authors present a novel fault-tolerant decentralized state estimator based on the polynomial chaos theory (PCT). This estimator has been developed to detect, diagnose, and isolate failed sensors. The sensor measurements are first diagnosed by the decentralized PCT sensor measurement estimators (DPCTSME), then are used to estimate the states of the non-linear electrical power system by means of a Decentralized PCT State Estimator. The use of PCT-based approach has been shown to be an effective tool for incorporating uncertainty in the estimation process. Numerical results of the application of this approach to an electric power test system are presented and commented.