Research on real-time admittance matrix identification based on WAMS and multiple linear regression

Abstract

This paper proposes the MLR (multiple linear regression) algorithm for parameter estimation of power system post-fault configuration based on the PMU real-time measurement and generator dynamic equations. Using real-time data collected by wide-area measurement system, the algorithm can take complicated cascading fault events into consideration without any information about the specific parameter and fault type or structure of the power system. This attractive feature avoids the difficult problem to determine the parameter and the topology state of a transient event in actual projects, making it possible for the complex perturbed trajectories prediction. The proposed algorithm has been tested on various sample power systems with promising and accurate simulation results.