Graduate student Ms. Leen Al Homoud presented a research paper entitled “Cyber-Physical Defense in Smart Distribution Networks” at the 53rd North American Power Symposium (NAPS 2021). Much of the existing literature has focused on mitigating data attacks resulting in nontechnical losses, false state estimation, and inaccurate load forecasting. However, cyber-physical security also needs to ensure that no grid operations are allowed that impact the physical operation of the system. For example, a malicious agent successfully hacks through the cyber security defenses of the grid, spoofs as a bona fide agent, and issues control commands that cause a voltage regulator to change tap position. While the actuation may not be severe enough to cause immediate damage to the system, it may causes accelerate wear-out of system components.
Along with her fellow authors Mr. Rinith Reghunath, Mr. Safin Bayes, Ph.D. student Aaqib Peerzada, Professor Katherine Davis and Professor Robert Balog, the paper investigated machine learning techniques that would detect anomalies in the control commands of an electrical power distribution system.