Non Gaussian State Estimation in Power Systems

Abstract

Most power system state estimators consider measurements as independent gaus-sian random variables and use a least square approach to estimate the most likely system state. In this paper we drop the questionable assumptions of independency and normality of measurements and develop techniques to accurately estimate the system state through appropriate transformations of correlated non-gaussian mea-surements. An example illustrates the proposed estimation technique. Conclusions are finally drawn.