Research On Vulnerability Mining Technology Of Industrial Control System For Intelligent Optimization

With the development of Internet technology,Industrial Internet has also been developed.The emergence of industrial Internet has greatly enhanced the automation of industrial production.However,because the environment of most industrial control systems is independent and closed,too much resources were not spent on safety issues at the beginning of the design.After connecting with the external network,although the industrial production efficiency has been improved,many safety problems have been exposed.This thesis introduces the architecture of the industrial control system,and mainly analyzes the vulnerability of the typical Modbus TCP protocol in the industrial control system.Through in-depth study of Peach Fuzzy Testing,Peach Pit test script is compiled to perform Fuzzy Testing on Modbus TCP protocol.Focusing on the path explosion and constraint solving problems in the process of Fuzzy Testing,the Attention LSTM model is established by combining Long-Short Term Memory(LSTM)with Attention,and an optimization method for Fuzz testing of Modbus TCP protocol based on Attention LSTM is proposed.The fuzzy data generated by Peach is used for packet capturing analysis,and the results are divided into Normal Case,Abnormal Case and Invalid Case,and the Attention LSTM model is used to judge whether the input use cases are invalid.Because the data set of industrial control system transmission protocol is time ordered,LSTM model is used to optimize the data set of Fuzz testing.After preprocessing and time sequence reconstruction,the fuzzy test data set is transmitted to the attention layer to calculate the attention weight at that time,and then the weighted sum operation is performed with the simultaneously engraved hidden state,the operation result is the attention distribution vector.After the attention distribution vector is obtained,the data sample features are obtained after the full connection layer.The final recognition results are obtained through the softmax classifier.The efficiency of fuzzy testing is improved by identifying and eliminating invalid cases.In order to verify the effectiveness of this method,Peach fuzzy test data set is used for verification.Modbus TCP protocol fuzzy test intelligent optimization method based on Attention LSTM is used to judge the identification accuracy and test case pass rate of invalid cases.The experimental results show that the identification accuracy of the Attention LSTM model for invalid cases reaches 97.03%,The pass rate of the test cases was82.15%,which realizes the high-precision identification of invalid cases and improves the pass rate of test cases.