Visualization and Classification of Power System Frequency Data Streams

Two challenges in the realization of the smart grid technology are the ability to visualize the deluge of expected data streams for global situational awareness; as well as the ability to detect disruptive and classify such events from spatially-distributed high-speed power system frequency measurements. This paper presents an interactive visualization model for high speed power system frequency data streams that displays both local and global views of the data streams for decision making process. It also presents a K-Median approach for clustering and identifying disruptive events in spatially-distributed data streams. The results from experimental evaluation on a variety of datasets show that K-Median achieve better performance and empowers analysts with the ability to make sense of a deluge of frequency measurements in a real-time situation.