Research On Delay Compensation And Schedule Algorithms Of Networked Control Systems

As the development of network technology and embedded systems, together with the increasing scale of control systems, there has appeared a kind of control systems which are based on communication network, viz. networked control systems (NCSs). The so-called networked control system is a real-time feedback control system which takes advantage of network to realize information interaction between distributed field sensors, controllers and actuators. It has many advantages compared with traditional control system, such as remote operation and control, facilitated installation and maintenance, more flexibility and reliability. But because of the introduction of communication network, there has emerged some new issues, which bring new challenges for analysis and design of control system, and it becomes the hot topic of control theory research.In this thesis, the research mainly focuses on network-induced delay compensation and network schedule in NCS.First of all, this thesis introduces the background, the construction and the features of NCS, and then the degradation for control system caused by network-induced delay is shown via theory analysis and simulation, after that some algorithms for compensating network-induced delay of NCS are introduced briefly.Secondly, the network-induced delay compensation algorithms based on uncertain parameters are studied. NCS with random network-induced delay which is shorter or longer than a period is considered respectively. And the time-delay discrete-time model with uncertain parameters for NCS is established in the two situations. Then the sufficient condition of system asymptotic stability is derived via Lyapunov stability theory in the form of linear matrix inequalities and corresponding state-feedback control law is designed. At last the numerical examples illustrate the effectiveness of the algorithms.Thirdly, Schedule algorithms for NCS are studied. NCS schedule platform including three control loops is established by TrueTime, and two schedule algorithms RM and EDF are compared in this platform. Then fuzzy schedule algorithm which is based on system error and error changing rate is proposed, and compared with EDF by simulation. The simulation result shows that fuzzy schedule algorithm can assign network resource more reasonable and make NCS get better overall performance when network resource is scarce.Finally, the semi-physical simulation platform which is totally separate from MATLAB is established by LabVIEW to make simulations more reality and more persuasion in this thesis. The network-induced delay compensation algorithm based on uncertain parameters is simulated in this platform using NEU Ethernet as Transmission medium. Network-induced delay is measured, and the effectiveness of the algorithm is verified in real network.