A tri-level optimization model to mitigate coordinated attacks on electric power systems in a cyber-physical environment

The electric power system is facing greater threats each year due to malicious attacks. In the recent decade, multiple attacks on both physical and cyber layers of the power system resulted in significant economic loss. In particular, a cyber-physical coordinated attack is deemed to be a destructive behavior that is likely to cause cascading failures. Therefore, this paper focuses on hedging against a coordinated cyber-physical attack from the standpoint of allocating defending resources. The coordinated attack involves physical short-circuiting of transmission lines after intruding the communication network of protection relays. The paper proposes a tri-level optimization model to formulate the coordinated attack scenario and identify the optimal defending strategy, which is a first of its kind study of defending resource allocation to hedge against the coordinated attack. Upper-level problem represents the actions of the planner and determines the optimal defending resource allocation on transmission lines and communication network of protection relays. Defending strategy achieves the lowest expected unserved energy with coordinated attacks considering multiple representative operating conditions. Middle-level problem formulates the behaviors of the attacker, who identifies the targets and attack time to maximize unserved energy. Lower-level problem simulates the actions of operator for system redispatch aimed at minimizing unserved energy. Sensitivity analyses are conducted on IEEE 14-bus system and IEEE 57-bus system to investigate the effects of attack budget, defend budget and restoration duration on unserved energy. The results indicate greater unserved energy with higher attack budget, lower defend budget, and longer restoration duration. Moreover, significant unserved energy reduction was observed when the proposed optimization model is applied to allocate defending resource.