Optimal Control Strategy Of Industrial Cyber-physical Systems Under DoS Attack

Industrial Cyber-Physical Systems(ICPS)is a multi-dimensional complex system integrating computing,communication and control.ICPS being integrated with Industrial Wireless Sensor Network(IWSN)has the advantages of resource sharing,flexibility,scalability and low maintenance cost.However,ICPS has shortcomings such as easy interference and easy attack due to the broadcast and open characteristics of industrial wireless sensor network information transmission,which seriously affects the stability and security of the system.The realization of ICPS modeling analysis,attack detection and compensation control under network attacks is the key and difficult point to ensure the stable and safe operation of the system.Therefore,this paper conducts research on DoS attacks and ICPS.From the perspective of achieving the security of ICPS under DoS attacks and reducing the impact of attacks on system stability,the following research work has been carried out:(1)Study the modeling problem of ICPS based on IWSN under DoS attack.To understand and analyze the dynamic behavior of ICPS,the ICPS under industrial wireless sensor network is modeled based on feedback control theory;the improved Bernoulli model is used to model the DoS attack,and the DoS attack is analyzed from the control perspective to the entire information physical system In order to formulate corresponding optimization control strategies.The establishment of a unified model provides theoretical basis and practical significance for the research and analysis of ICPS under DoS attacks.(2)Research the tracking control of ICPS based on periodic event triggering mechanism under DoS attack.In order to reduce the traffic of industrial wireless sensor networks and reduce the pressure on the network bandwidth,a periodic event trigger mechanism is designed to reduce the pressure on the network bandwidth and ensure the stability of the system.In order to realize the stable control of industrial cyber physical systems under certain performance index and make the system get rid of the limitation of equilibrium state,a state estimator and tracking regulator are designed,and then combine to get the LQG controller,to realize the tracking control of ICPS.(3)Research attack detection schemes and compensation strategies for DoS attacks in ICPS.In order to detect DoS attacks in the system,used a combined detection scheme which be the combination of Kalman filter and chi-square detector.The Kalman filter calculates the measurement residuals while filtering out the environmental noise.The chi-square detector normalizes the measurement residuals to obtain the detection value,and selects the detection threshold under the equivalence rate through training,and then combined with the established attack detection decision rules to determine whether the system is suffering from DoS attacks.In order to reduce the impact of the measured value of the attack moment on the controller after the detector detects the attack,a multi-step state estimation compensation strategy is designed to reduce the impact of DoS attacks on the system stability to a certain extent.In order to prove the effectiveness of the optimal control strategy adopted for ICPS,the ball-beam system was selected as the controlled object of ICPS,The Simulink/TrueTime platform was used to simulate the ball-beam system which based on wireless sensor networks.1)The DoS attack model based on the Bernoulli model can effectively destroy the stable application of the industrial information physics system under the feedback control model;2)Kalman filter can effectively suppress environmental noise in ICPS.Based on the separation law,the LGQ controller combined with Kalman filter and tracking regulator can achieve the tracking control of ICPS under certain performance indicators.The periodic event trigger mechanism can reduces the pressure on the communication bandwidth without affecting the stability of the system;3)The combined detection scheme of Kalman filter and chi-square detector is effective,and can detect DoS attacks better than the detection scheme based on Euclidean detector.The multi-step state estimation compensation strategy following the attack detecting can reduce the impact of the attack on the system to a certain extent.