Data Enhanced Robust Distribution System State Estimation for Realistic System With Parameter Corrections

Abstract

The accuracy of distribution line parameters significantly impacts distribution system state estimation (DSSE) since they are important components of measurement equations. This paper proposes a data-enhanced linearized power flow model for DSSE, in which the elements in the admittance matrix derived from distribution line parameters are corrected based on available measurements. Then, more accurate measurement functions can be established for DSSE. To deal with the occurrence of bad data, the Huber-estimator is used with the newly derived model for robust DSSE. One significant benefit of the proposed method is that the improved linear power flow model for measurement function construction guarantees convergence of the DSSE while the nonlinear AC power flow model may yield divergence issues in the presence of model parameter uncertainties. Morerover, the massive zero-injection information that exists in practical systems is utilized to enhance grid observability. Experimental results carried out on the realistic, unbalanced, three-phase, 2135-node Dominion Energy distribution system highlight the improved DSSE accuracy and convergence in the presence of model parameter uncertainties and bad data.