Event-Triggered Control Of Uncertain Nonlinear Systems Using Intermittent Output/State

Event triggered control has attracted more and more attention because it only sends the corresponding signal when certain clearly defined triggering conditions are met,avoids unnecessary signal transmission,can save communication resources and reduce the computational burden.The results show that compared with the traditional periodic sampling,event triggered sampling can improve the communication efficiency and ensure satisfactory stability and control performance,so it has become an alternative scheme of time triggered mechanism.Although the event based control law is similar to the classical continuous time sampling in form,the event triggered control problem brings many new challenges because only intermittent sampling information can be used in the control design.The research on event triggered control has made great progress,but the existing research results still have many limitations.It is mainly reflected in: 1)at present,the control strategy of introducing event triggering mechanism at the sensor side,and the system considered is generally simple(the high-order nonlinear system with mismatched parameter uncertainty has not been considered);2)For high-order nonlinear systems with mismatched parameter uncertainties,the current approaches only consider the event triggering mechanism at the controller side.For this kind of system,considering setting the event triggering mechanism in the dual channel of sensor-to-controller and controller-to-actuator,there is no relevant result at present.From the perspective of control,this paper studies the design of event triggered control for uncertain nonlinear systems with limited communication and computing resources.Because the intermittent state/output signal caused by the event triggering mechanism is non differentiable,the controller design and stability analysis become extremely challenging.This paper aims to solve the problem of how to use intermittent state/output feedback signal for control design under the framework of adaptive backstepping technology,and avoid direct derivation of intermittent state/output signal.The main research results can be summarized as follows:(1)For a class of nonlinear interconnected systems with dynamic input/output interaction,a state triggered control scheme is proposed.When the event-triggering mechanism is introduced to transmit the state signals of interconnected system,the virtual control signal involved in backstepping technique is non-differentiable,which makes the recursive backstepping design process unsuitable.To solve this problem,this paper proposes a set of replacement methods.Firstly,the continuous state feedback mechanism is used to design the decentralized adaptive controller,and then the intermittent state is used to replace the continuous state to design the actual event triggered controller,so as to avoid deriving the intermittent state signal directly,and solve the design problem of decentralized adaptive control under intermittent state feedback for the first time.(2)For a more general class of mismatched uncertain nonlinear systems,an output feedback event triggered control scheme using only intermittent output signal is proposed for the first time.Considering that both input and output signals are triggered,this paper designs a state observer using intermittent output signal,introduces the idea of dynamic filtering technique,and constructs a dynamic filter to avoid direct differentiation of virtual control signals in control design.In order to overcome the problem that the virtual control signal is still non differentiable in Lyapunov stability analysis,this paper first uses the original continuous output signal to design the auxiliary differentiable virtual control signal,and then uses the intermittent output signal to replace the continuous output signal to design the event-triggered controller.The results show that the proposed controller can ensure that all closed-loop signals are semi-globally bounded,and the transient performance in the sense of mean square error can be adjusted by properly selecting the design parameters.(3)For a class of mismatched uncertain nonlinear systems,an adaptive output feedback control scheme based on input and output triggering is proposed for the first time.This paper first uses the dynamic filtering technique to avoid the direct derivation of the virtual control signal in the backstepping design,then introduces the parameter projection technique into the design of the parameter estimator,and finally uses the triggered input and output signals to design an appropriate state observer to overcome the difficulty of unmeasurable state.In this paper,the projection operator is used in the adaptive law to ensure the boundedness of parameter estimation,and it is convenient to properly deal with the influence of error caused by output triggering.The results show that the adaptive backstepping controller proposed in this paper can ensure that all signals in the closed-loop system are semi-globally bounded,and the transient performance in the sense of mean square error can be adjusted by selecting appropriate design parameters.(4)For a class of nonlinear systems with mismatched uncertainty,an event triggered control scheme with possible sensor failure is designed for the first time.In this paper,an appropriate state observer is designed by using intermittent and faulty output signal,the virtual control signal is designed by using the estimated output signal,and the idea of dynamic filtering technique is adopted.It makes the process of control design and stability analysis easier,saves cost,and relaxes the constraints on unknown parameters.Theoretical analysis shows that through the sensor fault compensation control scheme based on output triggering proposed in this paper,even in the case of sensor fault,all signals of the closed-loop system are still bounded.