Font Size: a A A

Research On DoS Attack Detection And Security Control Scheme For A Class Of Linear Cyber-physical Systems

Posted on:2022-11-12Degree:MasterType:Thesis
Country:ChinaCandidate:X T WeiFull Text:PDF
GTID:2518306743972759Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
Cyber-Physical Systems(CPS)is a new generation of smart systems that combine computer,network and physical environment,it has significant strategic value and wide application potential.However,its complex and diverse system structures and the multiple applied fields can lead to vulnerability to the external attacks,thereby reducing the overall security of the system.The above problems have turn into a central issue in the field of CPS research.Moreover,the Denial of Service(DoS)attacks,as a more frequent attack type occured in CPS security threats,has also received widespread concern.CPS includes linear systems and non-linear systems.Among them,the non-strictly constrained non-linear CPS can be transformed into the linear CPS.In this paper,for the linear CPS with sensor DoS attacks,an adaptive sliding mode observer and high-order sliding mode differentiator are designed to realize attack detection and state estimation;for the linear CPS with actuator DoS attacks and unknown external disturbances,a security control scheme is introduced to offset the effects of attacks and disturbances on the system.These researches provide a feasibility research foundation and method for the nonlinear CPS researches.The work done in this paper is as follows:(1)Analyze the system structure and work process of CPS,and describe the physical process of CPS using differential equations(sets)in control system theory.By establishing a typical CPS non-attack model and a theoretical model of DoS attack itself,a generalized model of the linear CPS under DoS attack is obtained,which lays theoretical basis for CPS system attack study.(2)Aiming at the problem of linear CPS being attacked by sensor DoS attacks,based on the sliding mode theory,an adaptive sliding mode observer and high-order sliding mode differentiator are designed to transform the sensor DoS attack detection and state estimation problems into a robust stabilization problem of extended augmented system.Firstly,establish a CPS model under sensor DoS attack,and design an adaptive sliding mode observer to realize attack detection and state estimation;secondly,introduce a high-order sliding mode differentiator to reduce the state estimation error.The Lyapunov stability proof is performed for the sensor DoS attack detection and state estimation.And the numerical simulations prove the effectuality of the designed adaptive sliding mode observer and high-order sliding mode differentiator.(3)Aiming at the problem of the linear CPS attacked by actuator DoS attacks and unknown external disturbances,the optimal control theory,zero-sum game theory,adaptive dynamic programming(ADP)and neural network are integrated to design a security control soluton.Firstly,through the optimal control and the zero-sum game theory,the optimal control signal under ideal conditions is derived.Secondly,for the approximation of the optimal control signal,the value iterative ADP is used to transform the linear CPS actuator DoS attacks and unknown external disturbances into the dynamic approximation of the optimal control policy,and use the Critic-only neural network to approach the specific policy.A Lyapunov stability analysis is carried out on the stability of this security control scheme.Finally,a numerical simulation example is used to prove the effectuality of the proposed safety control scheme.
Keywords/Search Tags:Cyber-Physical Systems, Denial of Service, Attack Detection, Safely Control, Adaptive Dynamic Programming
PDF Full Text Request
Related items