Title of Submission
System Identification of Control Systems via Cyber-Physical Intelligence Attacks
Degree Program
Computing, PhD
Major Advisor Name
Hoda Mehrpouyan
Type of Submission
Scholarly Poster
Abstract
The smart revolution of control systems is characterized by communication networks. They provide more responsive, precise, reliable, and efficient systems with better operational and management capabilities. Despite these advantages, they also expose control systems to new threats which originate in the cyber domain.
Literature indicates that accurate system knowledge is necessary for attackers to design and carry out complex cyber-attacks. Cyber-physical intelligence (CPI) attacks are a state-of-the-art data-drive system identification methodology to infer system knowledge.
In this work, our CPI attack directly target the Proportional-Integral-Derivative gain values on the controller. Preliminary results demonstrate that accurate system identification is feasible.