Research On Quantized And Event-triggered Control

In the past few decades,with the rapid development of computer science,communi-cation technology and control theory,the scale,structure and complexity of the controlled system keep enlarging.Therefore,the traditional centralized control scenario cannot meet the modern control requirements because of its limitations in wire arrangement,equipment installment/maintenance,scale expansion,etc.In contrast,networked control systems(NC-Ss)are spatially distributed systems in which the communication between sensors,actua-tors,and controllers occurs through a shared bandlimited digital communication network.Since NCSs can reduce the complexity of systems with nominal economical investmen-t,ensure efficient and reliable information exchange,and adjust the number of sensors,actuators and controllers without heavy structural changes to the whole system,network control has been widely applied to practical systems.Meanwhile,NCSs have a high requirement to the bandwidth and transmission speed of the communication channels.Since a communication network is usually shared by dif-ferent system nodes while the network resources including communication channel band-width and computation abilities are limited,the signal blocking or data dropouts may occur which will degrade the system performance.Thus,how to address the problem of energy,computation and communication constraints is of great significance both in theory and in practice.Quantized control and event-triggered control are two most important techniques in decreasing the signal transmission burden of communication channels.However,quantized and event-triggered control also bring new challenges in designing suitable controllers,i.e.how to handle the measurement errors caused by signal quantization and/or event-triggering mechanisms.Note that the existing results of quantized/event-triggered control are still very limited,especially for uncertain nonlinear systems.In this paper,we aim to design new quantized and event-triggered control schemes for NCSs so that desired system performances can be obtained.The main work and contributions of this thesis are listed as follows:(1).The problem of adaptive output-feedback tracking control for a class of uncertain nonlinear systems with input quantization is studied.Firstly,a control scheme based on a newly-proposed quantizer is proposed.This quantizer is a combination of a logarithmic(or a hysteresis)quantizer and a uniform quantizer,and it has the advantages of both being logarithmic and uniform quantizers in ensuring reducible communication expenses and acceptable quantization errors for better system performances.It is shown that the designed adaptive controller ensures global boundedness of all the signals in the closed-loop system and enables the tracking error to exponentially converge towards a compact set which is adjustable.Then,a new robust adaptive controller is designed.This controller is more general as it can be applied to all the currently existing static quantizers.(2).The global stabilization problem for a class of uncertain nonlinear systems with unknown growth rate by output feedback is investigated.Both the output signal and the input signal of the system are quantized for the sake of less communication burden.A new control law with an adaptive gain is proposed to compensate for the quantization errors.It is proved that the proposed scheme ensures that all the closed-loop signals are globally bounded.In addition,the output signal can be regulated to a bounded compact set which is explicitly given.(3).The problem of event-triggerred adaptive control for a class of uncertain nonlinear systems is considered.The proposed control schemes remove the input-to-state stability(ISS)assumption with respect to the measurement errors and allow the system to contain unknown parameters.Three different event-triggering mechanisms are presented,i.e.the fixed threshold strategy,the relative threshold strategy and the switching threshold strategy.Moreover,rigorous stability proofs for all the three strategies are respectively given.(4).A new event-triggered control scheme for a class of uncertain nonlinear systems subject to actuator failures is designed.The actuator failure modes are allowed to be un-known and the total number of failures could be infinite.To reduce the communication burden from the controller to the actuator,a novel event-triggered control law is designed.It is proved through Lyapunov analyses that the proposed control protocol ensures that all the signals of the closed-loop system are globally bounded and the system output tracking error can exponentially converge to a residual which can be made arbitrarily small.