False Data Detection in Distributed Oscillation Mode Estimation using Hierarchical k-means

Wide-area oscillation monitoring and control in power systems is essential for preventing large-scale outages. To detect and estimate oscillation modes in real-time distributed measurement-based mode estimation approaches have been proposed. Unfortunately, these methods are vulnerable to false data injection attacks. In this paper we propose a hierarchical k-means false data detection approach to detect and remove false data from distributed Prony algorithms for oscillation mode estimation. Our proposed method is able to detect both multi-goal adversaries and noisy outliers equally well. We empirically illustrate the resiliency of our method against different attacks and show that it can detect corrupted data and accurately estimate oscillation modes.