Network Control System For Scheduling And Simulation

Networked Control Systems (NCSs) is a frontier research which combines network technology and control science. This thesis studies the scheduling and the simulation problems of networked control systems. The contributions of the works are listed in the following:First of all, this thesis analyses the characteristic, the advantages and disadvantages of four main types of control networks. Then, it introduces the technical characteristics and the methods of using TrueTime toolbox, which is the foundation of the NCSs simulation.Secondly, this thesis elaborates the concept of information scheduling in NCSs systematically, and divides the scheduling strategies of NCSs into three kinds: the static scheduling strategies, the dynamic scheduling strategies, and the mixed traffic scheduling strategies, and analyse their advantages and disadvantages. Then it studies three typical scheduling strategies: EDF, MEF, and MTS. Through simulation, which includes three kinds of information, it compares the characteristics and applicable scope of three scheduling methods.Based on the scheduling methods, this thesis proposes an intelligent dynamic scheduling based on Quality of Service (QoS) and Quality of Performace (QoP) of NCSs, by adjusting the sample periods of the control systems on-line. When the performance of system satisfy the requirements, time-varying sampling period scheduling can reduces the network resource of control information, and then the aperiodic information can get more network resource, obtain higher QoS under certain circumstances, thus the overall performance of NCSs can be improved.The simulation research indicates that, the dynamic scheduling strategy based on QoS and QoP can effectively improve the QoS of different kinds of information in NCSs, and optimize NCSs' overall performance.