Cyber attack and fault detection in DC microgrids by designing an event-triggered based-robust algorithm.

Area of Science:

• Electrical Engineering
• Cybersecurity
• Control Systems

Background:

• Large-scale systems (LSSs), particularly low-voltage direct current microgrids (DC MGs), are vulnerable to cyber-attacks.
• Existing detection methods may not adequately address complex attack vectors like false data injection and stealth attacks.

Purpose of the Study:

• To propose a novel distributed attack detection framework for LSSs, focusing on DC MGs.
• To enhance the resilience of DC MGs against sophisticated cyber-attacks through integrated detection modules.

Main Methods:

• Integration of an event-triggered (ET) observer for local monitoring and distributed unknown input (UI) observers for inter-subsystem assessment.
• Incorporation of an adaptive compensation mechanism for improved robustness against disturbances.
• Development of an ET control strategy for consensus, Zeno behavior prevention, and reduced communication overhead.

Main Results:

• The proposed framework effectively identifies attacks, including false data injection, stealth, and replay attacks, that standalone modules miss.
• Simulations demonstrate improved resilience against process noise and model uncertainties.
• The integrated detection units significantly enhance the overall security posture of DC MGs.

Conclusions:

• The novel distributed framework offers superior attack detection capabilities for DC MGs compared to traditional methods.
• The adaptive compensation and ET control strategy contribute to robust and efficient system operation.
• Future research should explore non-ideal communication and dynamic system topologies for enhanced autonomous control.