Abstract
This paper is dedicated to studying the voltage prevention control problem of DC microgrid clusters (MGCs) under scale random cyberattacks (SRCAs). Unlike the current mainstream modeling approach for random cyberattacks based on Bernoulli independent distributions, a random cyberattack model related to Bayesian distribution is proposed, which implies different probabilities of being attacked caused by the different network nodes in the communication network structure. Meanwhile, in the completely random sense, the energy level possessed by the attacker is reflected as the scale of system damage, which characterizes malicious attack behavior as the one-time random energy pouring problem. To cope with the established cyberattack model, we design an asynchronous federated intrusion detection scheme that matches the bilevel communication method adopted for the DC MGC in this paper. While ensuring the security of monitoring data, a recursive Bayesian analyzer is introduced to suppress the bias of prediction results caused by few inaccurate data. After considering the scaling features of random cyberattacks, a link reconstruction strategy is adopted instead of resilience control to ensure voltage restoration of large-scale clusters. For the sake of accelerating convergence speed, distributed finite-time control protocols are further designed, and the time response matching condition for achieving voltage synchronous restoration of DC MGCs under finite-time stability are obtained. Finally, the effectiveness of the proposed scheme is verified through real-time OPAL-RT simulation with the IEEE 123-bus distribution system.
Funding Information
  • National Science and Technology Major Project of China (2022ZD0116900)
  • National Natural Science Foundation of China (62273162)
  • Key Research and Development Project of Hubei Province of China (2025BAB010)
  • ‘the Fundamental Research Funds for the Central Universities’ (2025JYCXJJ070)

This publication has 19 references indexed in Scilit: