Research On Security Situation Awareness Of Industrial Control Network

Connecting the industrial control network to the Internet can not only greatly improve the working efficiency of the staff,but also make statistical induction of the production data,so as to better understand and control the industrial control system,forming a virtuous circle.However,if the industrial control network which is not originally connected to the Internet and has low network security protection is connected to the Internet,hackers can directly attack the industrial control network through the Internet.Therefore,using situation awareness technology to ensure the industrial control network security is a feasible method.Through situation awareness technology,the system can judge the status of the current industrial control network,and provide a reliable data for security personnel to predict the security situation of the industrial control network.Therefore,two key steps in the process of situation awareness,situation understanding and situation prediction,are studied by using artificial neural network algorithm.The specific work is as follows:Aiming at the situation understanding of industrial control network security,an intrusion detection method of industrial control system ARP(address resolution protocol)based on convolutional neural network and bi-directional short-term memory network hybrid model is studied: First,the convolution neural network is used to extract the data features,then the bidirectional short and long-term memory network is used to detect whether ARP attacks occur in the network according to the time sequence features of ARP messages in the network when attacked.Finally,the full connection network and softmax function are used to output the detection results.Using this intrusion detection method can detect whether the industrial control system is attacked by ARP more quickly and accurately,so as to better understand the situation of industrial control network.Aiming at the situation prediction of industrial control network security,a hybrid algorithm based on the combination of analytic hierarchy process(AHP)and BP(back propagation)neural network is studied to evaluate the risk of industrial control system.Security personnel can predict the security status of industrial control network in the future through the results of risk assessment.This risk assessment method first makes a more scientific and comprehensive initial assessment model by using the information security level assessment standard,then calculates the weight of each index of the model by AHP,selects the more important index according to the comprehensive weight as the input of BP neural network,and finally trains the neural network with the historical assessment data as the data set.This hybrid algorithm applies the advantages of AHP algorithm in screening the factors that affect the results,and combines the advantages of BP neural network in data fitting,better combines quantitative and qualitative analysis,makes the evaluation results more accurate,and reduces the differences of risk evaluation results of different teams for the same system,so that security personnel can better control the industrial control network To predict the safety status of.