Research On Whitelist Technology Of Industrial Firewall Based On GA-DE-SVM

Industrial firewall occupies an important position in the field of industrial control network security.Its core measure to achieve security protection is to use the characteristics of industrial control data analysis to build a whitelist database to prevent abnormal behavior.First of all,this article gives a certain introduction to the research status of industrial firewalls in industrial control systems,and then describes the network architecture of industrial control systems.At the same time,it specifically describes the Modbus TCP protocol,which is currently the most widely used in the Chinese market,and gives a detailed explanation of the pros and cons of the protocol itself.In particular,using the compatibility of industrial firewalls,whitelisting technology and classifiers are integrated to construct a discriminant model for illegal data.Secondly,the industrial control data sample is too large and has a large number of dimensions.When using the support vector machine to classify the data,it is prone to obvious defects.In order to make the classification effect of the support vector machine more prominent.Therefore,a whitelist technology using genetic algorithm(GA)and differential evolution algorithm(DE)to jointly optimize SVM is proposed.First,encode the obtained data,and combine the crossover and mutation links of GA and DE.The optimal SVM parameter penalty factor c and Gaussian kernel parameter g are selected through the combined GA-DE algorithm,and the algorithm is used for classification training.The simulation results show that the hybrid genetic differential evolution algorithm designed in this paper has a good effect on data classification.Finally,by building an automatic canning control system for analog conveyor belts,and using PLC to carry out industrial control experiments for simulation verification,the experimental results show that the algorithm has a certain improvement in its classification effect compared with a single SVM combined with particle swarms or genetic algorithms.Moreover,the test accuracy rate of normal data has also increased by 13%,while the test accuracy rate of abnormal data has increased by 10% compared with that of abnormal data,which verifies the effectiveness of the algorithm designed in this paper.