Machine Learning Techniques for Power Grid Protection and Control

Add to your list(s) Download to your calendar using vCal

• Prof Eduardo Cotilla-Sanchez, Oregon State University, USA
• Thursday 28 April 2022, 11:00-12:00
• Denis Gabor Seminar Room, 611.

If you have a question about this talk, please contact Fei Teng.

In order to make a power grid resilient to cyber or physical attacks, it is necessary to ensure that power system control and protection schemes can function properly to isolate faults, mitigate damage, and recover lost components, in the presence of an on-going attack. However, critical power system control and protection decisions such as microgrid islanding, relay tripping, and load shedding are inherently data-dependent, and they can be exploited by an adversary through data integrity attacks to produce malicious control decisions. In this talk we will focus on Phasor Measurement Data (PMU) applications, which are becoming increasingly important for power system operations, and can be falsified under various attack scenarios. There is a need for a countermeasure that can ensure PMU data integrity, is scalable, and is feasible for real time operation. Existing countermeasures are not sufficiently satisfying these requirements. We are working to fill this gap by developing and validating real time data correction and adversarial machine learning techniques for PMU data analytics. The second part of this talk focuses on the creation of an attack-resilient learning scheme for predicting the state of islanding or reconnecting microgrids. We build a classifier that uses machine learning techniques and PMU data that is resilient to cyber-attacks. The goal of this learning scheme is to be able to determine dynamically whether the reconnection of an islanded microgrid would lead to a stable or unstable network. It is important that the process is robust due to the potential of PMUs being compromised during the decision to reconnect or not. The proposed machine learning algorithm makes use of a small set of secure PMUs to achieve relatively accurate predictions for the stability of reconnecting islands.

Bio: Eduardo Cotilla-Sanchez is an Associate Professor of Electrical and Computer Engineering in the School of Electrical Engineering & Computer Science (EECS) at Oregon State University. He received M.S. and Ph.D. degrees in Electrical Engineering from the University of Vermont. Dr. Cotilla-Sanchez has over 12 years of experience in power system protection and reliability, with a focus on energy access. He has led several federally funded projects, including US Department of Energy (DOE) awards, and is a co-PI of the DOE Cyber Resilient Energy Delivery Consortium (CREDC). His grid integration work spans from rural minigrid distributed energy resources to transmission level renewable generation. His technical contributions are focused on power system modeling, resilience, and security. These interests spire into several other research areas such as nonlinear dynamics, complex systems, smart grids, microgrids, and wide-area power system data. Cotilla-Sanchez is the Vice-Chair of the IEEE Cascading Failures Working Group and also serves as Associate School Head for Graduate Programs in the School of EECS .

This talk is part of the Control and Power Seminars series.

This talk is included in these lists:

• Andrea Picciau's list
• Control and Power Seminars
• Denis Gabor Seminar Room, 611

Note that ex-directory lists are not shown.