Network-based Finite-time Control For Periodic Systems

Due to the prosperity and development of the Internet,abundant communication resources are provided to individuals,companies and governments.Meanwhile,the investigators in control field perceive the advantages of network and attempt to combine it with control systems,whose controllers,sensors and actuators exchange real-time information over a shared network.Therefore,the concept of NCSs is gradually proposed.In contrast to traditional point-to-point control system,NCS breaks function constraints,which result from the constraints of physical extention.Furthermore,it enables the integrated diagnostics and the execution of remote tasks to be more intelligent and convenient while ensuring the efficiency of systems.NCS is extraly favored by scientific researchers and engineers for its incomparable advantages and has been widely applied in various fields like aerospace,remote control and intelligent manufacturing.There is no denying that communications medium of network is not always reliable with some fundamental problems,which may result in poor performances or even lead to instability like time delay and packet dropout.As is known to us all,traditional Lyapunov asymptotic stability mainly concentrates on the steady-state performance in an infinite-time interval rather than transient-state performance.However,some bad transient performances maybe still exist,such as large overshoot which is not permitted to appear in industrial process when the system exists saturation.Conversely,finite-time stability(FTS)focuses on the transient performance and its state does not exceed the given bound.Therefore,the research on using finite-time control approach to deal with NCSs is urgently necessary.In this thesis,we focus on the issue of the finite-time control for periodic systems under several different environments of network.The main contributions are provided:1.The issue of FTS analysis and control for discrete-time periodic systems with random occurence of transmission delays in NCSs is investigated.We assume that the exogenous disturbance is l2norm-bounded and random transmission delays can be described as a Bernoulli process.Then,after setting a Lyapunov-Krasovskii functional,the sufficient conditions that can guarantee the system to be FTS and FTB are derived.After that the linear matrix inequalities(LMIs)technique is applying in a flexible way and a numerical case based on the previous conditions is presented to manifest the effectiveness of the design.2.The problem of FTS analysis and quantized output feedback control for discrete-time periodic systems with nonlinearities and randomly packet dropouts in NCSs is addressed.The nonlinearity of periodic system is expressed as the sector-bounded nonlinear function with the same period,the packet dropout process is modeled as a Bernoulli process with the hold-input strategy and logarithmic quantizer is chosen to quantize the measured output.A LyapunovKrasovskii functional is built and sufficient conditions are established to guarantee the control plant to be FTS and FTB respectively,then the finite-time method for the target system is designed via output feedback.Furthermore,we conduct an numerical example to validate the feasibility of the design.3.The topic of FTS analysis and output feedback control for discrete-time periodic systems with nonlinearities and deception attacks in NCSs is discussed.We choose the sector-bounded nonlinear to stimulate phenomenon of nonlinearity and consider both sensorto-controller and the controller-to-actuator deception attacks are modeled as two independent Bernoulli processes,besides,we assume that there exists an upper bound of the injected signals.The Lyapunov-Krasovskii functional which is based on the above assumptions is used to analyze the sufficient conditions for FTS and FTB of known system,after that a controller is presented in a set of LMIs.A numerical example is provided at the end of chapter,which can prove the effect of method.