Testing equality constraint hypotheses in weighted least squares state estimators

Abstract

The presence of a gross error in analog, status, or parameter data can dramatically affect both the convergence and the quality of the results yielded by weighted least squares state estimators. Most of the literature available on the subject deals with analog errors. This paper extends the well known J-test to the validation of hypothesis represented as linear equality constraints. The proposed testing approach has been applied to validate hypothesis about the status of switching devices (open/closed) in networks modeled at the physical level (bus-section/switching-device model).