Research On Model Predictive Control For Networked Systems With Time Delay

With the development of computer,communication and control technologies,the networked control systems(NCSs),which exchange the information between each node of the system through a network,have been developed rapidly and been widely used in many fields.NCSs have many advantages by comparing with the tradition control system,such as less wiring,lower cost,easier maintenance and expanding.Due to these distinctive benefits,NCSs have also attracted much more attention by many experts and scholars and become a hot research in the field of control theory in recent years.However,some new problems have been brought up due to the application of network in the real-time control system,such as network-induced delay,packet disordering,packet dropout and so on.If ignoring these problems during the model of the system or the design of the controller,it may affect the control performance of system,even resulted in instability of the system.In this thesis,the model predictive control for NCSs with time delay and packet disordering is investigated.The main research works are listed as follows:1.The design of stabilizing predictive controller problem is studied for NCSs with random short delay and input constraints.Based on enhancing the read frequency of the buffer,the closed-loop system is described as a Markov jump system.Sufficient conditions for the closed-loop NCSs to be stochastically stable are derived by MPC method.Furthermore,the state feedback control laws are obtained by solving a convex optimization problem with linear matrix inequalities(LMIs)constraints.Finally,the effectiveness of the proposed method is verified by simulation.(?).The packet reordering and controller design problem are studied for NCSs with random long delay.Firstly,we present a model of selecting of the latest data packet by the actuator to deal with packet disordering,and the resulted NCSs are modeled as Markov jump systems.The calculation method of the reordered transition probability matrix is given.Furthermore,based on the Lyapunov stability theory,the sufficient conditions for stochastic stability of the closed-loop NCSs are derived,and the predicting controller is solved by LMI.Finally,the effectiveness of the proposed method is verified by simulation.3.The design of stabilizing controller based dynamic matrix control is studied for NCSs with time-varying delay.Through the timing analysis,a new method is proposed to deal with packet disordering at the controller and the actuator,and the resulted NCSs model is established.Then,a modified dynamic matrix control algorithm is proposed based on the established model,and the design procedure of the controller is also given.Furthermore,sufficient conditions to ensure the stability of the closed-loop system are derived.Finally,the validity of the proposed method is verified by simulation.