This is the third in a series of profiles and updates from the U.S. Department of Energy’s (DOE) first Electricity Industry Technology and Practices Innovation Challenge (EITPIC).
The EITPIC was designed to tap into American ingenuity for ideas on how to make the nation’s electric grid stronger and more resilient. Through the challenge, DOE sought ideas from industry, academia, and other innovators for technologies and solutions to address existing or emerging vulnerabilities and threats to the electric sector or mitigate interdependencies between the electricity sector and other sectors.
Washington State University Vancouver received $25,000 for its proposal to develop a data processing user interface that would increase the quality of data analytics for the electric utility sector. The proliferation of data from phasor measurement units (PMUs) has opened new opportunities to develop tools for grid system monitoring and operations support, including situational awareness. However, the diversity of data processing systems increases costs and lowers reliability. The team’s proposal would increase the performance, scalability, and quality of services (QoS) of PMU data analytics through the development of a novel data processing system, called UPS, for supporting QoS-aware data-driven synchrophasor workflows.
“A variety of PMU data processing systems have been proposed to improve the reliability, security, and efficiency of the nation’s electric grid,” said Xuechen Zhang, Assistant Professor in the School of Engineering and Computer Science at Washington State University Vancouver. “However, the disparities between the diverse data processing systems lead to a computing environment that not only drastically increase the system management costs but also limits the performance, scalability, and quality of services (QoS) of PMU data analytics.”
Three novel contributions were proposed by the team that, if successful, would develop the first data-traffic aware workflow management system; devise a unified input-output layer to support data consolidation with predictable performance for system operators; and create a system that consolidates data and avoids large-scale data movement between storage clusters.
The EITPIC award will support the design of a new PMU data-driven workflow management system in collaboration with a team of interdisciplinary researchers with expertise in computer science and power system management. Next, the team will profile the compute-bound PMU data analytics and construct machine learning models for the prediction of their execution time online.