Research On Output Feedback Networked Predictive Control Based On Interval Type-2 T-S Fuzzy Model

Under the background of deep integration of informatization and industrialization,it is an inevitable trend to apply network technology to transform traditional industry and realize networked control of industrial processes.However,it also brings great challenges to MPC.The traditional MPC cannot deal with the problems of quantization and packet loss that signals must encounter in the transmission process.If the industrial process using traditional MPC is directly placed in the network environment,it will have a negative impact on the system control performance and safety production.Hence,aiming to solve the problems of quantization and packet loss in networked control systems(NCSs),this thesis systematically studies the synthesis approaches of output feedback MPC based on interval type-2(IT2)Takagi-Sugeno(T-S)fuzzy model,by using the Lyapunov stability theory as the basic method to guarantee the system performance,and taking the quadratic boundedness technique and linear matrix inequality(LMI)method as the basic tools.The main contents of this thesis are divided into three aspects:1.The synthesis approach of output feedback MPC for IT2 T-S fuzzy NCSs with packet loss and quantization is studied.Firstly,the sector bound approach is used to deal with the quantization error,a Markov chain is introduced to characterize the process of random packet loss,and a mathematical model of nonlinear systems with data quantization and packet loss is established.Then,based on the offline designed state observer,the MPC optimization problem which minimizes the upper bound of the infinite horizon performance objective function is online solved.Finally,according to the obtained estimated state,a strategy of refreshing estimation error bound which can guarantee the recursive feasibility of optimization problem is introduced.2.For the IT2 T-S fuzzy NCSs with bounded disturbance,the synthesis approaches of output feedback MPC without free control move and with free control move are investigated,respectively.Firstly,a method of describing the stability for NCSs with bounded disturbance is obtained by introducing the quadratic boundedness technique.Based on the provided method,the LMI conditions of the state observer are given.Then,by parameterizing the infinite horizon performance objective function into a single state feedback law(a free control move and a single state feedback law),the synthesis approaches of output feedback MPC without free control move(with free control move)are proposed under bounded disturbance.Finally,the recursive feasibility of these two approaches is analyzed respectively,and the proof of closed-loop stability is given.3.The synthesis approach of event-triggered dynamic output feedback MPC for IT2 T-S fuzzy NCSs with bounded disturbance and packet loss is studied.Firstly,in order to reduce the communication burden,an event-triggered scheme is designed to determine whether the current measurement output should be released into network communication links.Then,by online minimizing the upper bound of the quadratic performance objective function,the synthesis approach of parameter dependent dynamic output feedback MPC can be obtained based on the quadratic boundedness technique.Finally,the auxiliary optimization problem of calculating a tighter estimation error bound is presented to improve the control performance.