Cyber-Resilient Self-Triggered Distributed Control of Networked Microgrids Against Multi-Layer DoS Attacks (2104.09250v2)

Abstract: Networked microgrids with high penetration of distributed generators have ubiquitous remote information exchange, which may be exposed to various cyber security threats. This paper, for the first time, addresses a consensus problem in terms of frequency synchronisation in networked microgrids subject to multi-layer denial of service (DoS) attacks, which could simultaneously affect communication, measurement and control actuation channels. A unified notion of Persistency-of-Data-Flow (PoDF) is proposed to characterise the data unavailability in different information network links, and further quantifies the multi-layer DoS effects on the hierarchical system. With PoDF, we provide a sufficient condition of the DoS attacks under which the consensus can be preserved with the proposed edge-based self-triggered distributed control framework. In addition, to mitigate the conservativeness of offline design against the worst-case attack across all agents, an online self-adaptive scheme of the control parameters is developed to fully utilise the latest available information of all data transmission channels. Finally, the effectiveness of the proposed cyber-resilient self-triggered distributed control is verified by representative case studies.