Resilient Control Of Cyber-physical System Under False Data Injection Attacks

The Cyber-physical System(abbreviated as CPS)refers to a type of new intelligent complex system which was highly integrated by interactive multi-dimensional computing units and physical systems in the network environment.The current research and assessment of CPS security is divided into information security and control security.Control security issues of CPS will be emphatically discussed later.As the most secretive and so phisticated attacks,False Data Injection(FDI)attacks can be implemented by hijacking physical devices(sensors,controllers,etc.)or network channels,Injecting false data or changing the signal transmission strategy increased the network delay and cha nged the signal transmission path without being detected by the system's detection mechanism,resulting in the system behavior being abnormal or destabilizing;it will also cause huge damage to the system;the main contents of this thesis as follows:First of all,the CPS will be mainly analyzed under the FDI attacks.To reduce the adverse effects on the CPS which under the FDI attacks,the structure and algorithm of the control system need to maximize the “resilient” of the control system which integrated by personnel and control devices.Here,“ resilient ” refers to the recovery speed of a system that can be restored to a normal state of operation after being destroyed or a certain level of security under slight degradation of performance.Based on the related literature that has been read,the definition of resilient control,how to quantify the performance of resilient control and the fields of its applications will be introduced in detail,at the same time,this thesis explains the differences and linkages between the resilient control and the existing control technology.Finally,the problems in the resilient control and its follow-up development are explained.Secondly,the CPS discrete linear time-invariant dynamic system model—output tracking control system is established.The Kalman filter is used to predict and correct the data transmitted by the system.The Euclidean-based detector calculates the observed data and the estimated data.The deviation between the values was used to detect FDI attacks and linear quadratic Gaussian(LQG)controllers.The FDI attack sequence were designed which will successfully evade the system monitoring and unfired the detector alarms and proved by simulation examples.Finally,on the basis of the above research,combi ned with flexible control and stability analysis of the system,the control system under FDI attack s and security augmentation system matrix are designed to reduce the impact of FDI attacks on system operation and the system's attenuation under FDI attacks.The degree of performance degradation improves the control performance of CPS when attacked by this kind of FDI attacks,uses a simulation example to verify its effectiveness.