Research On Elastic Control Of Industrial Cyber-Physical Systems Under Network Attacks

Industrial-Cyber-Physical-Systems(ICPS)is a close connection between the Industrial Wireless Sensor Networks(IWSN)and the traditional Industrial Control Systems(ICS),where the computing and communication elements cooperate to control physical objects.However,due to the open environment of the wireless communication network,ICPS is susceptible to various types of network attacks causing data loss,theft,and tampering in the transmission and communication process,which can further cause physical malfunction,system performance being significantly decreased,and stability being destroyed.In order to meet the security requirements of maintaining good performance under ICPS network attacks,the key and difficulty is construct an effective defense strategy based on the traditional cyber physical systems model for achieving the ICPS flexible strategy.Therefore,starting from the perspective of an attacker,this paper studies the flexible control strategy of the industrial cyber physical systems of linear time-invariant discrete systems with aiming at reducing the damage of the physical system caused by attacks.(1)Research on the behavior characteristics of industrial cyber physical systems attackers,and make qualitative and quantitative analysis on the performance and attack strategies of three typical network attacks: denial-of-service attack,replay attack,and false data injection attack.Network attacks can affect the physical process by feedback driven.Therefore,in the analysis framework of the model,mathematical methods are used to quantitatively describe attacks,and a unified system behavior model that expresses the physical damage caused by attacks is established,to build a foundation for constructing effective defense measures.(2)Research on Network attacks in industrial cyber physical systems,especially the problem of data injection and system performance damage caused by the attack.By introducing the robust control theory,the physical impact caused by network attacks is regarded as a fault signal,and the structural model of industrial cyber physical systems is improved.The state estimator is designed to obtain the attack time,and the residual estimator is used as an input to design the fault estimator to quantify the physical impact of the attack and compensate for the attack,the fault-tolerant controller is further designed to ensure that the closed-loop system can maintain the mean square index stability and meet flexible requirements under attack.(3)Research on the attacks defense scheme of industrial cyber physical systems under the cooperation of multiple sensor nodes.A distributed filtering scheme is designed for the set of sensor nodes,and the information of the sensors and neighboring nodes is reasonably used to accurately capture the status of each node.At the same time,the random switching process of the sensor network topology and the data packet loss caused by random are considered,the attack estimation and dynamic output tracking feedback controller are designed for each node,which further improves the system flexibility under the uncertain network environment.Simulink / TrueTime / LMI was used as a simulation tool to analyze and verify the validity of the research method.The simulation results show that: 1)The attack space model and the attacker attack model studied according to the nature of different attacks can effectively describe the attack characteristics of Network attacks;2)Attack compensation and output feedback control scheme can greatly reduce the impact of attacks in the sensor channel on the basis of maintaining the stability of the system;3)The distributed filtering scheme based on multi-sensor nodes can accurately assess the status of each node,and the attack compensation and tracking feedback controller based on distributed filtering can alleviate the damage to the system while tracking a given signal.