Oscillation monitoring from ambient PMU measurements by Frequency Domain Decomposition

This paper extends the method of Frequency Domain Decomposition (FDD) towards real-time analysis of ambient PMU measurements in power systems for the purpose of oscillation monitoring. The main idea of FDD is to apply Singular Value Decomposition (SVD) to the power density spectrum matrix. The resulting singular values correspond to individual modes under specific conditions. The damping ratio, modal frequency and the mode shape of poorly damped (with less than 5% damping ratios) oscillatory modes can be directly determined from ambient PMU measurements. The theoretical background is presented in the paper. The technique is tested on small linear systems to show its effectiveness and weaknesses. Finally, the application of FDD in power system is shown.