Robustness Analysis of Power System considering Incubation Period for LLM Training Requirement

With the deep integration of intelligent technology, the risk of cyber attacks on cyber physical power systems is also increasing. Previous studies have primarily focused on the impact of network structure and virus transmission mechanisms on CPPS robustness, often neglecting the actual incubation period, which significantly influences the difficulty and efficiency of launching cyber attacks. In this paper, we examine the impact of the incubation period on CPPS robustness by analyzing the effects of the overload threshold, considering various coupling patterns between power and cyber networks. Furthermore, to identify the most vulnerable nodes, we employ an adaptive control strategy—the optimal load shedding model. Simulation results reveal that a higher overload threshold can enhance robustness, especially in scenarios where the virus incubation period is short. After the incubation period, attackers may target critical nodes to maximize attack effectiveness and reduce CPPS robustness, rather than compromising all infected nodes, particularly when the incubation period is prolonged. Our work offers cost-effective insights for improving CPPS robustness.