Model Predictive Control Of Networked Cascade Control System

Cascade control system wherein the control loops are closed through a real-time network is called a networked cascade control system (NCCS for short). NCCS is a special kind of networked control system, which fully combines the advantages of both cascade control system and networked control system. It not only can reduce system costs, improve system diagnostic maintenance level, but also can quickly overcome the influence of secondary disturbance and improve the system efficiency. The introduction of network brings us convenience; it also increases the complicacy of analyses and designs on control system, and makes many problems for the system. The obvious problems are the network-induced delay effect on the system, and the data packet dropout phenomena owing to the uncertainty of the network transmission. The existence of network-induced delay and data packet dropout phenomena will result in the NCCS control performance degradation and even destabilize the system. Considered the problems of network-induced delay and data packet dripout in NCCS, the model predictive control algorithm of networked cascade control system is researched in this thesis.To networked cascade control system for the study, the backgrounds and significances, basic problems and the current situation development of NCCS are discussed at first. Additionally, according to different configurations and the locations of network, four structures of NCCS are introduced, and take the type II NCCS based on fieldbus for study, modeling and analysis the type II NCCS with network-induced delay and data packet dropout respectively.For the phenomenon of network-induced delay and data packet dropout, model predictive control of NCCS are studied in this paper. Based on modeling NCCS with data packet dropout as Markov link, state feedback model predictive control algorithm of NCCS with data packet dropout is proposed which can guarantee the closed-loop NCCS asymptotical stability and minimize receding horizon performance index is considered. However, in the actual industrial production process, the system states are not fully measured, we only consider output feedback control. Thus, in the case of network-induced delay is less than a sampling period; output feedback model predictive control algorithm of NCCS with network-induced delay is studied.The model predictive control of networked cascade control system inherits many advantages of predictive control; the simulations results illustrate the effectiveness and feasibility of the proposed method.