Research On Risk Assessment Strategy For Power Cyber-Physical System

The deep integration of traditional power system and information and communication technology has formed the Cyber Physical System(CPS).The deep integration of the information system and the power grid physical system improves the observable controllability.It also exacerbates the risk of cybersecurity threats affecting the normal operation of the physical grid across space.The cyber security threats faced by power CPS are becoming more and more serious,and it is necessary to strengthen the research on risk assessment methods of power cyber-physical fusion systems.Accurately assessing system risks and taking effective risk prevention and control measures can improve the safe operation level of power CPS.Therefore,this paper studies the security risk analysis of power CPS under network intrusion attacks,and proposes a quantitative assessment method for power CPS risk.This paper firstly analyzes the research status of power CPS on fusion system modeling,network attack risk assessment methods and network attack defense methods.At the same time,the model framework and network security threats of electric power CPS,as well as the information-physical cross-layer risk propagation model,are analyzed,which provide the basis and direction for the subsequent proposed electric power CPS risk assessment method.Secondly,from the attacker’s point of view,a power CPS risk assessment method is proposed based on the best attack path of the network.The evaluation method firstly establishes the network attack model of electric power CPS,and then studies the method to find the best attack path of the network.And the traditional Q-learning algorithm randomly selects attack actions,which leads to the slow convergence of the algorithm,which makes the algorithm’s exploration process of the best attack path of the network easier to converge.Then,the system risk assessment method and quantitative index of power CPS under network attack are studied.And the simulation of IEEE 14 power distribution system proves the effectiveness of the improved algorithm in this paper to explore the best attack path of the network.At the same time,the feasibility of the power CPS risk assessment method based on the best network attack path proposed in this paper is proved.Finally,combined with the network attack and its defense strategy,the multi-agent game model is studied.Based on this model,the network threat propagation process of power CPS is given,and the risk assessment method of power CPS based on the multi-agent game model is proposed.This method firstly establishes the multi-agent game model of electric CPS under network attack,taking the network attacker and the electric CPS defender as the target of opposing multi-agents in the game,and the electric CPS system as the interactive environment of the two game parties.The network attacker selects the network attack action with the probability distribution strategy,and the power CPS defender deploys the defense resources with the probability distribution strategy,forming a two-person matrix game,and finally transforming the calculation of the Nash equilibrium value into the target optimization problem solved by the particle swarm algorithm.Combined with the Nash equilibrium strategy,the risk assessment method of power CPS under the attack and defense game is proposed,and an example is used to prove that an effective defense strategy can improve the safety of power CPS operation,and has a relatively good effect on reducing system risk.