Ensuring Resilience in Networked DC Microgrid Systems: Leveraging Dynamic Signature Analysis for Intrusion Detection and Response

This study presents an active defense mechanism aimed at identifying cyber-attacks on an Islanded DC microgrid utilizing distributed consensus control. The proposed mechanism involves injecting a unique random signal, termed as a "watermark," into the control inputs of the system. This watermark signal is demonstrated to propagate through the actuator, converter, and microgrid components, thus becoming evident in all measurement data used for system control. The proposed method uses the gathered data to create a system identification model, which includes a constant system matrix for representing the system in state-space form, enabling accurate estimation of future states.Under normal operating conditions, the properties of the watermark signal remain unchanged, while they are altered during attack scenarios. Subsequently, two statistical sequence tests are employed to accurately detect cyber-attacks. This approach effectively detects any malicious manipulation of measurement data, such as false data injection or record and replay attacks, perpetrated by malicious actors seeking to gain control of the microgrid or disrupt its operations, potentially leading to significant grid damage.The robustness of the proposed active defense mechanism is verified through validation with various attack scenarios in the MATLAB environment.