Research On Design Of Attack Parameter Dependent Controller For Networked Linear Stochastic Control Systems Under Network Attacks

Stochastic systems are a class of systems that describe actual operation influenced by random factors such as external environmental disturbances and internal structural changes,such as aircraft systems and intelligent manufacturing systems,and have important practical applications.With the rapid development of networked control systems(NCSs),networked stochastic control systems(NSCSs)have also received more and more attention.In this context,cyber attacks have become a challenge that must be faced to protect the security of NSCSs.Therefore,it is of great importance to solve the security problem of NSCSs under network attacks.Based on the analysis of the above-mentioned status quo,this paper makes a series of researches on how to carry out security control of networked linear stochastic control system under the influence of network attacks,and the following research results have been achieved.Firstly,the problem of attack parameter-dependent controller design for networked linear stochastic control systems under periodic denial-of-service(Do S)attacks is investigated.A switched stochastic system model is developed under the impact of the Do S attacks.The stability of the switched stochastic system is analysed by introducing a Do S attack parameter-dependent time-varying stochastic Lyapunov function.A mean square exponential stability criterion for the stochastic system is given,while enabling the closed-loop system to have L2-gain performance.Based on this,a design scheme for an attack parameter dependent controller is given based on the linear matrix inequalities(LMIs)technique.Finally,the effectiveness of the scheme is verified with an aircraft system as an example.Secondly,the problem of co-design of attack parameter-dependent controllers and observers for networked linear stochastic control systems under periodic Do S attacks is investigated.The Do S attack is modelled as a periodic pulse-width modulated interference signal,and a Luenberger observer is constructed to estimate the unmeasurable system state.An observer-based controller is designed and a new model of the switched stochastic system is derived.Then,a Do S attack parameter-dependent time-varying stochastic Lyapunov function is introduced to analyse the switched stochastic system.The criterion of mean square exponential stability of the switched stochastic system is given.A co-design scheme of attack parameter-dependent controller and observer is also given by applying LMIs technique.Finally,the effectiveness of the scheme is verified with an aircraft system as an example.Finally,the problem of observer-based attack parameter-dependent controller design for networked linear stochastic control systems under periodic Do S attacks and False data injection(FDI)attacks is investigated.The idea of a piecewise extended observer-based controller to handle Do S attacks and FDI attacks is presented.This simplifies the modeling of error systems for stochastic systems with unpredictable states,thus reducing the difficulty of observer-based resilient controller design.The attack parameter-dependent time-varying stochastic Lyapunov function is introduced to analyze the mean-square exponential stability of the error system,and a co-design scheme of piecewise extended observer gain and L2 controller gain is proposed.Finally,the effectiveness of the scheme is verified with an aircraft system as an example.