Research On Modeling And Robust Control Problem Of Networked Control Systems

This kind of control systems wherein the control loops are closed through a real-time network can be called networked control systems (NCSs). Networked control system is new hot-topic in control engineering. Since uncertain delays, network fault and disturbance are widely known to degrade the performance of a networked control system, at the same time, networked control system is an integrated research area, which is not only concerned about control, but also relevant to communication network. So we must combine the knowledge of control and communication network together to improve the system performance. Following this direction, it is worth researching robust controller design combined with system modeling. In this dissertation, we have deep research on modeling and robust controller design for networked control system. We design some novel models and accordingly introduce some robust controller and transport rules, which can robustly preserve the performance of networked control system in terms of uncertain delays, network fault and disturbance.In this dissertation, the present research situation of networked control system is extensively introduced. On the basis of analyzing the present models and robust control design method of networked control system, the idea of solving the performance degradation caused by uncertain delay, network fault and disturbance is also presented.Because the controller can only receive the delayed state signal, a new delayed state-variable model for networked control systems is presented. Based on which, the model of networked control system with uncertainty and disturbance is also designed.For the design of robust transport protocol, a kind of switched model of networked control system is introduced. According to the time-variant delay, a fuzzy logic switched control model is also given.A new delayed model for networked control system is presented, based on which a LQR controller is designed. A method of delays-estimation online is also discussed. For the difficulty on implementation of LQR in NCS with time-variant delays, a fuzzy logic with LQR controller is introdued.Based on the networked control system with multiple delays and uncertainty, some control rules are designed. Considering the network delay, environment and model uncertainty, a sufficient existence condition of robust guaranteed cost controller is discussed. A method of LMI design is also given. Considering the disturbance, a sufficient existence condition of r suboptimal H∞controller is presented. A method of LMI design is also introduced.Based on the new delayed state-variable model, a model of sensor fault is given for networked control systems. The theory of robust fault tolerant controller design is introdued. The method of LMI design based on MATLAB is also given. The condition of actuator fault is considered too.The reduction of bandwidth necessitated by the communication network in a networked control system is considered. A method on reducing the number of data packet exchanging between the sensor and the controller/actuator is explored. A novel structure of controller and special transmission rule are designed for reducing the number of data packet and making the system stable.The modeling and robust controller design is an extensive research topic. therefore there is a lot of work to do in this direction. Include that sufficiently consider the time-variant delay, network fault and external disturbance in modeling of networked control system. More novel hybrid model can be design for networked control system. New fuzzy-neural network control rule can be considered for networked control system. New delay-dependent robust controller can be presented for networked control system. New robust transport control protocol can be design for networked control system. At last, apply the research results to practice.