Game theory-based vulnerability analysis of cyber-physical power systems under interdependent network attacks

Cyber-physical power systems (CPPSs) have gained widespread adoption worldwide, driven by the growing need for enhanced power system security. The convergence of digital technologies with traditional power systems has facilitated significant enhancements in system monitoring, control, and power transmission. However, due to the interdependence of cyber-physical networks, CPPSs significantly increase their exposure to security risks under coordinated cyber-physical network attacks that can simultaneously compromise the transmission lines and communication links. In this paper, we develop a bilevel game theory-based attack-defence (GTAD) model that considers both functional and topological interdependence between the cyber and physical networks to assess the vulnerability of CPPSs under coordinated cyber-physical network attacks. Specifically, the upper-level problem aims to maximise load loss in the power system through coordinated cyber-physical network attacks; the lower-level problem aims to minimise load loss to the power system through corrective generation dispatch actions. Then, a strong duality-based method and a big-M method are employed to reformulate the GTAD model into a single-level mixed-integer linear programming (MILP) model. Finally, case studies are conducted on the IEEE RTS 24-bus system and a practical 36-zone Great Britain power transmission system to demonstrate the validity and rationality of our proposed GTAD method in determining higher risk and more coordinated cyber-physical network attack strategies.