Rotor Speed Estimation of Synchronous Generators Based on PMU Data and Tuning of Linearized Swing Equation Coefficients

Abstract

The rotor speed of Synchronous Generators (SGs) is a fundamental dynamic state that reflects the stored kinetic energy and the generator’s inertial response to system disturbances. Accurate rotor-speed estimation enables the detection of poorly damped electromechanical oscillations that may lead to instability. However, precise assessment of SG dynamics is often hindered by model inaccuracies and parametric uncertainties. With the increasing deployment of Phasor Measurement Units (PMUs), novel approaches have emerged to enhance post-event system diagnosis, particularly in low-inertia systems. This paper proposes a two-stage framework for estimating the rotor speed of SGs by tuning the coefficients of the linearized swing equation using data from PMUs, thus eliminating the need for the complete Electric Power System (EPS) model. In the first stage, the Morris sensitivity method identifies the most influential coefficients to refine the parameter-estimation domain. In the second stage, an Entropy-Based Genetic Algorithm (EB-GA) refines these coefficients by minimizing the mean squared error between measured EPS dynamics and the model response. Simulation results on the IEEE 68-bus system demonstrate that the proposed framework effectively captures SG rotor-speed dynamics, even under small perturbations and noisy PMU data. The EB-GA achieves more precise parameter tuning than conventional genetic algorithms, improving the representation of SG dynamics for post-event analysis and subsequent controller adjustment in EPSs.