Driving along the highway, most people have seen parts of the U.S. power grid in the form of tall, metal towers carrying power lines. These power lines are transmitting much of the electricity we use daily to power microwaves, refrigerators, washers, dryers and more.
But with this vast amount of electricity and power, many components aren’t yet accounted for — including techniques for analyzing data, specifically the methods used to measure data. These techniques are critical to preventing error, faults or hacking within the power grid.
In the Department of Electrical and Computer Engineering (ECpE) at Iowa State University, Professors Venkatramana Ajjarapu (PI), Manimaran Govindarasu, and Umesh Vaidya recently received a $1 million grant from the National Science Foundation (NSF) for predicting power grid behavior during potential disturbances and protecting the stability of power grid systems in the future through mathematical algorithms for measuring, analyzing and mapping grid data. The data analytics resulting from this project will allow researchers and scientists to have data-driven mapping for power grids.
By measuring critical system parameters within the power grid, estimation of grid condition can be more accurate. With high fidelity real-time measurements, researchers and scientists can more accurately predict the reasons behind faults in power grid systems.
In the present state-of-the-art, data-driven dynamical system analysis, the time evolution of the underlying dynamic states is required. Getting all of these dynamical states measurements in the real world is impossible or expensive, and mostly a subset of output measurements are available. In the power system scenario, the Phasor Measurement Units (PMUs) and the micro-PMUs (μ-PMUs) measure voltage and current phasors with high granularity, thus capturing the system dynamics.
“In the future, if there’s an outage, can we explain why? There are so many components — anything can go wrong,” Ajjarapu said. “This project will allow us to predict power system behavior during disturbances and take proactive control actions to avoid or mitigate undesirable behavior.”
In this project, the team is proposing to use these measurements to perform near real-time applications for various dynamic events that occur in the power systems.
Causality has been a topic of research in the fields of philosophy, economics, biology and networks to understand the cause-effect relation in a given system. Identifying causal interaction can enable researchers to understand the causation of dynamic phenomena and their impact on system components.
“We aim to develop a new measure of information transfer to characterize causal interactions in a dynamical system. The proposed notion of information transfer will be realized in a purely data-driven scenario,” Ajjarapu said.
The critical measurements immediately after a disturbance contain the information about the disturbance and how it can affect other components in the system.
“We are trying to identify this interaction through casualty mapping. The challenge here is to quickly calculate this information in real time to take control actions to prevent undesirable aspects of the disturbance,” Ajjarapu said.
The team is also developing an integrated tool to emulate the power system with communication networks that can be used for the sensor measurements and control commands.
“The tool will integrate communication networks to the power grid simulation framework to form the integrated power-cyber co-simulator (IPCC). The IPCC will be used for synthetic data generation that will emulate the real-world cyber-physical power system, and this data will be used for the development and validation of the proposed data analytics. We will validate the proposed data analytics techniques in two ways: using the IPCC framework to test the accuracy of the proposed methods, and using ISU’s hardware-in-the-loop (HIL) cyber-physical testbed (PowerCyber Testbed) to validate the real-time applications,” Ajjarapu said.
Ajjarapu has spent a large amount of time in his career analyzing power grids and devise methods to mitigate fault events. Vaidya has done extensive research in dynamical systems and its applications to power systems. Govindarasu brings expertise in cyber-physical system security and HIL testbed-based experimentation for power grids. A lot of their studies tie into analyzing, mapping and protecting power grids — because a power grid simply has so many levels of intricacy.
“I was very excited when I found out about this award, especially given that the project will leverage interdisciplinary expertise and realistic datasets,” Ajjarapu said. “The thing here is, we are already moving forward in this direction, with many other components involved, so with this award they are all coming together.”
Power grids are complex. But using this team’s data-driven techniques, researchers may be one step closer to understanding, predicting and protecting power grid systems.