State Estimation And Fault Detection For Networked Control Systems Under DoS Jamming Attacks

In recent years,networked control systems are bound to become the trend in the development of artificial intelligence automation due to the rapid development and integration of information technology,digital technology,control technology,communication engineering and computer science,and are widely used in modern control areas,such as power grids and power systems,intelligent transportation systems,health care and chemical production systems.Networked control systems have many unique advantages such as high flexibility,low cost and resources sharing,but the factors that control networks grew in size,more network nodes,the influence of sudden disturb fault and the network's openness bring about some new problems which cannot be ignored.The problem of security control caused by network attacks becomes one of the most noteworthy research hotspots.DoS(Denial of service)jamming attacks is a kind of resource-exhausted attack with high frequency and seriously attacks influence,which would affect the quality of network services.In the practicalities of industrial production,due to the states of many systems cannot be measured directly or difficult to be measured,observer could be used to estimate the unknown state of the system.The traditional transmission mode in networked control systems is the time-triggered mechanism,which would inevitably cause the unnecessary waste of network resources.Therefore,it is necessary to introduce event triggering strategy to realize the distribution according to needs in order to save bandwidth and improve the efficiency of resources usage in networked control systems.Based on the above analysis,this paper investigates the state estimation and fault detection of networked control systems in the presence of DoS jamming attacks,including the follows:Firstly,consider the networked control systems subjected to periodic DoS jamming attacks,and take restrictions on this attacks signals and modeling.Then,an observer-based event-triggered control scheme is proposed by using the switched observer method,and the exponential stability of system is also obtained by utilizing time varying piecewise Lyapunov function plus free weighting matrix method.And then,co-design observer and resilient event-triggered controller.Secondly,study the networked control systems subjected to unknown period DoS jamming attacks,and make some assumptions and constraints on the attacks and modeling.Then the output-based event triggering mechanism and the analysis to stability of the system are received,and the sufficient condition for the existence of the controller is also received by using the linear matrix inequality technique.Based on the above condition,the co-design of observer gain matrix,output-based event-triggering parameters and resilient controller is obtained.Finally,the design of fault detection filter for linear stochastic systems under external disturbances,fault signals and DoS jamming attacks is researched.A new resilient event-triggered strategy is proposed,which could resist DoS attacks and save bandwidth and other network resources.The fault detection residual model is constructed and then its stability and H_? performance are analyzed.Last,the co-design of the resilient event-triggered and fault detection filter is acquired by linear matrix inequality technique.