Identification of coherent trajectories by modal characteristics and hierarchical agglomerative clustering

This paper introduces a novel method to identify coherent generators using the inter-area modal characteristics of power systems. The key idea is to extract the inter-area modes from the simulated data and then to apply a clustering strategy. Thus, the proposed method consists of extracting the phase of the oscillatory modes via a modal identification technique and applying a hierarchical agglomerative clustering technique together with the Elbow's method to gather the phases of each mode, enabling to provide coherent trajectories of generators. The proposed method uses a Taylor-Fourier filtering strategy to remove noises and nonlinearity in the time evolution of coherent generators. Simulated signals with noise added are used for assessing the proposition. Results corroborate the proposed strategy for identifying coherent trajectories in large-scale power systems.