Linear Power Grid State Estimation with Modeling Uncertainties

TL;DR

This paper introduces a linear probabilistic power grid state estimation method that effectively incorporates measurement and model uncertainties, demonstrated on a large-scale transmission network with accurate results.

Contribution

A novel probabilistic framework with a new RTU model that makes power system state estimation linear without sacrificing accuracy.

Findings

Accurately estimates system state distributions including true states.

Mean of estimated distributions matches traditional deterministic estimates.

Scalable to large transmission networks like the entire USA.

Abstract

Recent advances in power system State Estimation (SE) have included equivalent circuit models for representing measurement data that allows incorporation of both PMU and RTU measurements within the state estimator. In this paper, we introduce a probabilistic framework with a new RTU model that renders the complete SE problem linear while not affecting its accuracy. It is demonstrated that the probabilistic state of a system can be efficiently and accurately estimated not only with the uncertainties from the measurement data, but also while including variations from transmission network models. To demonstrate accuracy and scalability we present probabilistic state estimation results for the 82k test case that represents the transmission level grid of the entire USA. It is shown that the estimated state distributions include the true grid state, while their mean exactly corresponds to the estimated deterministic state obtained from the nonlinear state estimator.