Research On Key Technologies Of Power Industrial Control Network Security Risk Analysis And Prediction

Electric power industry is the basic industry of national economy.The importance of electric power secure production is decided by the electric power production,the electric power manufacture construction,the objective law and production characteristic of the electric power diversified operation and social function.If the power supply is interrupted,especially when power outages in large areas caused by power grid accidents happen,the production of all walks of life will be forced to halt,bringing chaos to the economy,society and people’s living order.What’s worse,it may cause social disasters,and result in extremely bad political impact.On the basis of analyzing and summarizing the existing active security defense techniques,the paper first studies the risk monitoring techniques based on log analysis and realizes the safety of industrial control system function.Second,the paper studies the techniques of vulnerability assessment and attack&defense diagram,and realizes the information security of the industrial control system which integrates detection,monitoring and auditing.Third,the paper researches on the security risk prediction and privacy protection techniques,and provides integrated defense solution which is rooted in the whole lifecycle of industrial control system,and perfectly integrates function security and information security.In summary,the research work and innovation of this paper include the following aspects:(1)In the field equipment layer,aiming to make the field equipment trusted,controllable,secure and efficient,we implement the bypass deployment monitoring system,and design the log analysis architecture based on the var:ious types of electric power industrial control system log data,and implement it on Big Data platform.According to the tiny change trend of the statistic characteristics of the flow attributes in the abnormal network traffic,this paper presents an ensemble prediction algorithm based on time series(EPABT),to partially resolve the problem of system anomaly quantity prediction based on time series.Meanwhile,to evaluate the accuracy of the algorithm,a new evaluation metric,i.e.,the asymmetric error cost function(AEC)is proposed aiming at the characteristics of power industrial control system,which further synthetically improves the internal security protection capability of the industrial control system.(2)In the process monitoring layer,aiming at the deficiency of the EPABT algorithm which only implements system monitoring locally,this paper proposes an overall game attack and defense security strategy based on the improved system vulnerability point scoring standard(UIVSS),which carries the professional and upgradeable security vulnerability database.It uses the phased game theory and considers the related factors,including the reality of game between the attack and defense sides,the cumulative successful attack probability,and the asset value of the attacked objects.The security risk control decision-making mechanism of the power industrial control system is put forward as a whole,to strengthen the network behavior supervision and promote the running of normal business in the smart grid.(3)In the business management layer,aiming at that basic safety equipment and safety products can not embody the characteristics of active defense to the most extent,with the massive historical data accumulated from the practice in electric power industry,a prediction update algorithm is proposed based on Classification Restricted Boltzmann Machine and Markov time-varying Model.The algorithm solves the problem of the low efficiency of traditional Boltzmann machine,and combines the Markov time-varying model to predict and analyze five different network states,which provides the decision making basis for the active security defense of the system.(4)In order to solve the problem of privacy leakage in the above stages,the paper implements differential privacy protection in frequent pattern mining based on the micro-aggregation algorithm,which ensures the balance among information release,data analysis requirements and privacy protection demands.It guarantees the trust of all parties in the power industrial control system and the healthy growth of power industrial control system.