Research On Industrial Internet Vulnerability And Traffic Monitoring Technology Based On CBAM And AFS

With the ever closer connection between industrial production and networks,industrial control production is becoming increasingly intelligent and the security of the industrial Internet is getting more and more attention from researchers.So far,there are a lot of installed industrial controls connected to industrial Internet,the private protocols in the equipment are numerous,and the industrial production environment is also closed,these characteristics make the industrial Internet security under certain threat.Due to the outdated equipment and systems used by the industrial Internet and the large amount of data,the ongoing process of program software is vulnerable to security vulnerabilities and abnormal traffic,posing a threat to normal production.So as to ensure the security of the industrial Internet,the thesis discusses the similarities and differences between traditional control networks and industrial control networks,improves the traditional network security model,and proposes an industrial Internet security monitoring model based on vulnerabilities and abnormal traffic.The main content and research results of the thesis are as follows:First of all,the industrial Internet security detection model based on vulnerabilities and abnormal traffic proposed in the thesis not only scans the source code before the program software runs to detect vulnerability information,but also collects data information after the program software runs to monitor traffic,forming a dual prevention mechanism.In addition,the security monitoring model built in the thesis can monitor the entire operation process of the industrial Internet,find and deal with the factors that endanger the safety of the industrial control network as soon as possible,and form a closed-loop industrial Internet security model to ensure the safe operation of the whole system.Secondly,the thesis proposes a method to detect the vulnerability of the industrial Internet based on CBAM-CNN-SVM.This method scans the source code before the program runs,builds the industrial Internet security vulnerability library through the existing open data set,processes the vulnerability code into a syntax abstract tree,and carries out numerical and normalization processing on it,and then inputs it into the convolutional neural network model combined with the two-channel attention mechanism for training.In the classification process,the last layer of the convolutional neural network is changed into a supporting vector machine with improved classification capability improves the accuracy of detecting industrial Internet security vulnerabilities.Finally,the thesis proposes an abnormal traffic monitoring method for industrial internet based on an improved K-means and AFS system.This method monitors traffic data after program software is run on the industrial internet,,uses KPCA method to extract features and reduce dimensions,and then uses the improved K-means algorithm,namely K-means++algorithm,combined with the artificial fish swarm algorithm to cluster the abnormal traffic.Analyze which traffic data is abnormal so that the network management personnel of the industrial control system can correct it in time and meet the security requirements of the industrial Internet.The simulation results show that the vulnerability detection and abnormal traffic monitoring methods used in the industrial internet security monitoring model based on vulnerabilities and abnormal traffic can accurately and comprehensively monitor whether there are security threats in the industrial internet,and have relatively complete closed-loop protection.At the same time,it improves the security performance of the industrial Internet security monitoring model,meets the requirements for real-time monitoring in the industrial Internet,ensures the security of the industrial control network system,and maintains the security of the industrial control system.