Data Integrity and Cyberattack Detection using Dynamic Watermarking for Resilient Microgrids

Cyber-attacks on microgrid systems, especially data manipulation attacks such as replay attack and Denial-of-Service (DoS), causes communication delay and unstable responses. Even though control strategies such as Consensus Control (CC) are able to coordinate electric current and voltage flow, they are at risk of malicious attacks. Communication delay leads to undetected changes in line current, and voltage leads to incorrect responses from the consensus controller, which overloads the microgrid in milliseconds. To address these challenges, this paper presents an Observer System (OS) based Dynamic Watermark (DW) detection model that detects delay-induced cyber-attacks during steady states and load fluctuations. We have developed a Grid-Specific Dynamic Watermarking (GSDW) signal that enhances real-time detection capabilities, resulting in a real-time non-zero residual showing cyber attack dynamics in the proposed observer system. Our detailed case study demonstrates real-time attack detection and prevention, ensuring the stability and integrity of Microgrid (MG) systems under challenging cyber threat conditions. Comprehensive simulations and validation demonstrate the practicality and efficacy of our approach in mitigating risks posed by delay-induced cyber attacks in MG systems.