Networked control systems(NCSs)are a class of advanced control systems in which communication networks are introduced in traditional control systems to implement the information interactions among system components,and the NCSs have several advantages such as low cost,easy maintenance,information sharing and remote control.NCSs have already been applied in many modern industrial control systems.Meanwhile,the networkinduced issues(such as network-induced delays and packet dropouts)directly affect the system stability and performance,and bring new challenges in analysis and synthesis of the NCSs.Some traditional control theory(such as Proportional-Integral(PI)control,and fuzzy control)may no longer fit the NCSs,which should be deeply reconsidered.Taking networkinduced delays and packet dropouts into consideration,this thesis investigates three problems using state augmentation methods:(i)the network-based PI control for a continuoustime system and its application in the vehicle hight adjusting of direct-drive-wheel(DDW)systems in a wireless network environment;(ii)the networked output tracking control for a linear continuous-time system with constant reference input and time-varying sampling and its application in speed tracking of DDW systems;(iii)the networked fuzzy static output feedback control for a class of nonlinear discrete-time systems with sensor saturation and measurement noise.The main contents of this thesis are listed as follows:(i)Considering both network-induced delays and stochastic packet dropouts and using two different artificial delayed states to characterize the update of proportional and integral control signals,respectively,based on a state augmentation method,the networkbased PI control system is modeled as a stochastic impulsive time delay system,which consists of a continuous-time system with two input delays and a reset equation.For the NCS,a novel Lyapunov-Krasovskii functional(LKF)with discontinuous terms is constructed by using the information on the two input delays and the upper bound of network-induced delays.Some less conservatism conditions for exponentially mean square stability and H_? performance are derived in terms of linear matrix inequalities(LMIs),respectively.A particle swarm optimization algorithm is presented to design the corresponding PI control gain.A Zig Bee-based simulation platform for networkbased PI control is built.The feasibility and effectiveness of the proposed results are verified by the platform and the Matlab Simulink,respectively.(ii)Based on the equilibriums of tracking error systems solved by using traditional PI control method and referring to the results in(i),considering the time-varying sampling and network-induced delay,the network-based PI controller is built and the closedloop tracking error system with two input delays is obtained using a state augmentation method.The output tracking problem is converted into an asymptotic stability problem of the closed-loop system.A delay-dependent asymptotic stability condition is established by using a discontinuous LKF method.Taking the DDW system as a case study,a PI controller design method is provided in terms of LMIs and the effectiveness of the proposed method is illustrated by simulation and experimental results.(iii)Inspired by the idea in(i)and(ii)that the PI control problem is converted into the static output feedback control problem,the networked fuzzy static output feedback control for discrete-time nonlinear systems is investigated,where the effects of the nonlinearity,sensor saturation and measurement noise in practical systems are considered simultaneously.The nonlinear NCS is modeled as an asynchronous discrete-time Takagi-Sugeno(T-S)fuzzy system subject to both interval-like input delays and sector nonlinearity constraints.By introducing a fuzzy LKF using a state augmentation method and employing the matrix-based quadratic convex method,an interval-delay-dependent bounded real lemma is derived such that the closed-loop system has a prescribed H_? performance.A fuzzy controller design method is presented by using a cone complementarity linearization algorithm.An inverted pendulum system via networked fuzzy control is provided to show that the proposed stability condition is of less conservatism and the control design method is effective. |