A Model-Data-Combined Identification Approach for Distribution Grid Line Parameter

Abstract

Compared with the transmission grid, the distribution grid has the characteristics of huge node scale, sparse measuring equipment and three-phase load imbalance. The parameter identification methods for the transmission grid might not suitable to the distribution network. In this paper, a two-stage model-data-combined identification approach for distribution grid line parameter is proposed. First, the linear initial identification model of line parameters is constructed based on the phase angle assumption, which is solved by the least square method. On the second stage, the refined identification model is developed based on the trust region reflection method which uses the initial values provided by the first stage. Meanwhile, the voltage phase angle is recovered. Finally, IEEE 33-bus and 74-bus distribution grid is utilized to verify the proposed method.