Research On Power Information Network Security Situation Awareness Under Intrusion Attack

With the rapid development of information and communication technology(ICT),the construction of power system information network is becoming more and more mature.Through the interaction of computing,communication and physical environment of power system,a multidimensional heterogeneous complex system with real-time perception,dynamic control and information service integration has been formed.At the same time,the complexity of information networks has made network security problems increasingly serious,and various network threats present higher demands on the security performance of power systems.How to accurately perceive the risk state and evaluation of power system information network has become one of the important tasks of power system network security research.In this paper,the situation awareness of power information network and the knowledge of risk assessment under attack are analyzed and studied.The main work includes:(1)Aiming at the threat of power information network attack,a situation awareness method based on LDA-RBF was proposed and applied to power information network security.Firstly,the method optimizes the sample data of power information network by using linear discriminant analysis(LDA),extracts the feature index effectively and obtains the best projection,which makes the sample have the best separability.Then the preprocessed data are used as training data of RBF neural network to realize security situational awareness of power information network and identify intrusion attacks in power information network.(2)Using the KDD Cup99 dataset and the cyber attack data in the power information network environment to simulate the proposed method.The results show that the network security situation awareness method based on LDA-RBF can achieve network security situational awareness more accurately and efficiently.(3)Studying the cyber attack in the environment of cyber physical fusion,analysing the network attack scenarios.On the basis of graph theory,the intrusion attack process associated with information device vulnerabilities is analyzed and modeled based on attack graph.Then bayesian inference method is applied to analyze the attack source information network transmission,and the intrusion attack is quantitatively evaluated by probability.The comprehensive load shedding or transient power angle stability margin is the risk value of the attack on the power system.