Robust Safeguarding of Frequency Regulation and Market Operations Via SVDD and Feature Engineering

The power grid's cyber-security is becoming more important as our reliance on real-time interaction and communication technology continues to expand. Any malfunction in these technologies might cause the automated generation control (AGC) system to fail entirely since they are essential to the system's functioning. Cyberattacks, if successful, may halt power grid operations and cause major financial and personal data breaches. If an attacker wants to cause power outages, they may, for instance, use false data injection (FDI) attacks to tamper with the communication channel's network settings and send out incorrect dispatches. This introduces a novel market-operation assault template that makes use of exponential functions to enable stealthy attackers to obtain bigger cash settlements. To effectively identify FDIs aimed against the AGC system, it presents a new intrusion detection approach based on SVDD. This method employs a number of learning strategies, such as K-means clustering, SMOTE, and a number of Support Vector Data Description models serving as foundational learners. The detection technique is enhanced to capture varied patterns during standard and endangered grid states by the introduction of a new feature variable. Additionally, a novel market operation assault based on foreign direct investment (FDI) is shown in the chapter. This attack maintains frequency grid checks while unjustly increasing the output of particular GENCOs for financial gain.