PROBABILISTIC STATE ESTIMATION IN DISTRIBUTION NETWORKS

The traditional approach to state estimation only provides the estimated values tothe network operator without any information about the accuracy of the estimates. If a network is made observable by using a large number of pseudo‐measurements, the estimated state may not be accurate enough to be practical even if the network is classified as observable. This represents a limitation of the traditional approach to observability. This chapter presents a new probabilistic approach to state estimation in distribution networks based on confidence. This probabilistic approach is then also extended to observability, resulting in a new observability assessment that focuses on the accuracy of the state estimation results. It focuses on the confidence that the estimated parameters are within their respective constraints. The observability analysis is performed before the state estimation and determines whether it is possible to estimate the network state with the available set of measurements or not.