Mining Spatial Frequency Time Series Data for Event Detection in Power Systems
Swati A. Lavand, Gopal R Gajjar, Shreevardhan Arunchandra Soman, Rajeev Kumar Gajbhiye · 13th International Conference on Development in Power System Protection 2016 (DPSP) · 2016
In this paper we address the question of automatic event detection and classification. The method proposed can be applied to both, real time streaming data and data stored in a historian. The scheme can classify the type of disturbance as well as identify the time instance of the disturbance. Thus, it will improve the post disturbance analysis. Frequency is a common signal in power system which responds to major power system disturbance like faults, load and generation tripping. A desirable feature of this signal is that under normal condition it is stationary. As such we propose a two step event detection scheme using time series of wide area frequency measurements e.g. from PMUs. We present case studies using historian PMU data from different locations.