False Data Topology Attack Modeling And Key Branch Identification In Power System

The two major blackouts in Ukraine in 2015 and 2016 showed that attackers were able to launch simultaneous,multi-stage,and multi-site cyberattacks by invading power information systems and performing long-term reconnaissance operations to obtaining the necessary information of power system.As the most common form of cyberattacks,false data injection attacks can invade the information network and tamper with system measurement and operation data in real time on the basis of bypassing bad data detection,mislead the state estimation results and interfer with the decision-making of the control center,causing system economic losses and even leading to power outages.At present,a relatively extensive exploration has been carried out on the model of false data injection attacks and the corresponding defense strategies.However,the existing false data injection attacks need to tamper with the bus load data collaboratively,resulting in the construction of attack vector being restricted by the load prediction and further affecting the attack effect.To overcome the above problems,this paper proposes the false topology attacks that cooperate to tamper with the branch power measurement data and protection information.The false topology attacks can be launched sneakingly without tampering with the bus load,it can bypass the bad data detection and get rid of the load prediction constraints to a certain extent.Therefore,this paper analyzes the load shedding of the power system under the attack by modeling the false topology attack,and identifies the key defense branches in the network,aiming to provide a new method for improving the systems’ ability to respond to the false topology attacks.The specific research contents of this paper are as follows:1.By analyzing the configuration of branch protection and bus protection in the actual system,this paper constructs an attacker-operator bi-level game optimization model that considers the cooperative tampering of branch measurement data and protection information,and convert it into a single-layer model according to KKT conditions to solve.The results show that the false topology attacks can cause system load shedding through faking break branches,and compared with the traditional load redistribution attack,it has the characteristics of strong concealment and high threat.2.Aiming at the AC state estimation model applied in the actual power system,a mathematical method is proposed to quickly convert the DC attack vector to the AC attack vector.At the same time,the feasibility of false topology attackers launching attacks without complete system information is verified.3.To identify the key defense branches of the system,by introducing defenders,a trilevel game optimization model of defender-attacker-operator is constructed,and the C&CG algorithm is exploited to realize the conversion and solution of the model.The simulation results show that by defending the key branches of the system,the load shedding of the system can be effectively reduced,and the ability of the system to respond to false topology attacks can be improved.4.To solve the deficiency of high time complexity in solving three-layer game optimization model,the key branch identification problem is converted into a Markov decision process,and the reinforcement learning algorithm is exploited to solve the model in real time.At the same time,to identify the key branches from the perspective of system economic operation and safe operation,a parallel search strategy for agents based on the safety-economic double Q-value function is proposed.Furthermore,to realize the online identification of key branches,an identification method against operating state based on relative entropy is proposed.By comparing the operating states in different times,the transfer strategy of the agent between different states is determined.The simulation results show that the online identification of defense branches can be realized by exploiting reinforcement learning algorithm.The key branch online identification method proposed in this paper provides a theoretical basis and new ideas for the power system to respond to the threat of the false topology attacks.