Detection And Identification Of Cyber-Attacks In Cyber-Physcial System And Its Applications

In recent years,Cyber-physical System(CPS)have been applied more and more widely in people’s life with the rapid development of Internet of Things technology.However,CPS may suffer from malicious cyber-attacks due to its inherent vulnerability,which causes inestimable losses.Therefore,CPS security issues are increasingly prominent and widely concerned by people.Detection and identification of cyber-attacks in CPS is a hot topic in the research of CPS security,but the current works on detection and identification of cyber-attacks in CPS usually only consider CPS of a single node.With the development of information technology and embedded technology,CPS may contain multiple nodes or even it is a large-scale coupled CPS such as smart grid and intelligent transportation.This paper studies the problem of detection and identification of cyber-attacks in CPS with several nodes,proposes effective detection and identification of cyber-attacks algorithms,and applies them to a DC microgrid model.And further,the effectiveness of the proposed algorithms is verified by theoretical analysis and experimental simulation.Specifically,CPS is a network coupled system that is coupled with multiple nodes and interfered by noises in this paper,among which the noises are bounded.By modeling two types of network attacks from network layer and physical layer,the corresponding network layer and physical layer attack detection and identification problems are proposed and studied.The main works of this paper are as follows:(1)Aiming at the problem of detection and identification of cyber-attacks on network layer,relevant technologies in cryptography are used to design a set of effective detection and identification of cyber-attacks algorithm on network layer in CPS.In particular,the sender will first embed the data to be transmitted wirelessly into the hash value,which is generated by the Logistics chaos system and Keccak algorithm.Then the modified data will be encrypted by the SM4 algorithm before transmission.After decryption and separation,the receiver will compare the hash value with the initial hash value.If there is any change,it indicates that cyber-attacks are launched from the network layer during the wireless transmission of the node and resulting in the tampering of the transmitted data.Thus,the detection and identification of cyber-attacks on network layer in CPS is realized,and the effectiveness of the proposed algorithm is verified by experimental simulation.(2)Aiming at the problem of detection and identification of cyber-attacks on physical layer,state estimation and other related technologies are used to design an effective detection and identification of cyber-attacks algorithm on physical layer in CPS.In particular,each node first uses an unknown input observer to realize a completely decentralized state estimation process.Then,the relevant residual signal sequence is calculated.Finally,the residual signal sequence will be input into the trained long short-term memory neural network and the output of whether the node is attacked on the physical layer can be obtained,so as to realize the detection and identification of cyber-attacks on physical layer in CPS,and the effectiveness of the algorithm is verified by experimental simulation.(3)In this paper,the proposed detection and identification of cyber-attacks algorithms in CPS are applied to the DC microgrid model,and the practicability and effectiveness of the algorithms are illustrated by experimental simulation.