Risk-based contingency analysis for power systems considering a combination of different types of cyber-attacks

Substation with various communication and control devices has become a major target of cyber-attacks. In general, two main types of cyber-attacks can be performed on substations through direct intelligent electronic device connections from remote accesses or via compromising local substation supervisory control center and data acquisition systems, thus respectively resulting in different damage to the power systems. In this paper, the combination of two types of cyber-attacks on the power systems are considered. The generalized stochastic Petri net is utilized to simulate dynamic intrusion processes of these two types of cyber-attacks. Furthermore, the successful attack probabilities of cyber-attacks are calculated based on simulation results. Then, a probabilistic bi-level optimization model is proposed to identify critical components with the maximum cyber risk and the types of cyber-attacks. The two types of cyber-attacks initiated from substation space to individual components are particularly modeled in the proposed model. Finally, a two-stage algorithm using the network flow is pro-posed to solve the proposed model more efficiently. Simulation results tested on the IEEE RTS 24-bus and the 118-bus systems validate the effectiveness of the proposed model. Results also show that the devised algorithm can improve the computation performance in dealing with larger-scale power systems.