Impulsive Modeling And Predictive Control Of The Networked Control Systems

With the fast development of computer technology, communication technology, network technology and control theory etc, the networked control technology and the networked control systems constructed upon network platform have been a hot research issue in the field of automation technology since the 1990s of the 20th century. This type of system has the advantage of greater flexibility with respect to traditional control systems. In addition, it allows for reduced wiring, as well as a lower installation cost. It also permits greater agility in diagnosis and maintenance procedures. Therefore, some problem such as time delay, data package loss, may occur in the network transmission, which can make the analysis to NCSs a difficult task. Because the classical control theory cannot be directly applied to NCSs'controller design. This dessertation is mainly composed as fellows:This paper makes use of impulse control theory to model the networked control systems into the network impulse-control system. It firstly introduces the model of impulsive dynamic system, and then it puts forward the hypothesis that there is no time delay and data packet dropout in the network and the networked control systems is modeled into the network impulse-control system, which has SISO and variable-sampling. According to the nonlinearity of the controlled quantity and controlled object, it provides two impulse models of nonlinear networked control systems, the two models are simplified into impulse models of linear networked control systems.When this paper discusses the networked control systems with time delay and data packet dropout, the system's sampling period is supposed to be time-variant, which effectively solves the problem of data packet dropout. As long as the interval between the two adjacent data packets when they reach actuator node is within the range of the maximum sampling internal allowed by the system, the system is stable. Based on this, by using the theory of delay impulse control system, the impulse model of nonlinear-networked control system with time delay is established, and then, the impulse model of linear networked control system with time delay is also established.We discuss the stability conditions about impulsive model of the networked control systems. Firstly, we suppose the samples set is known in advance, and then, we introduce the model of impulsive dynamical systems, and establish exponential stability of nonlinear time-varying impulsive systems by employing multiple Lyapunov functions with discontinuities at a countable set of times. Our stability and stabilization results are presented as Linear Matrix Inequalities (LMIs). By solving these LMIs, one can find an upper bound on the sampling intervals such that the stability of the closed-loop is guaranteed. Secondly, we present exponential stability theorems for delay impulsive systems by employing multiple Lyapunov functional with discontinuities at a countable set of times. By solving appropriate stability LMIs, we can find the relationship between the max sampling interval and the delay for which the stability of the closed-loop system is guaranteed.We introduce the conversion method between the NCSs'impulse model and the CARIMA. In this thesis, a new-networked control strategy is developed which takes advantage of the potential capability of Generalized Predictive Control to compensate for anticipated data transmission problems. Considering the prediction steps can not make a great fluctuations, while, the predict step include the delay's fluctuations within the maximum range. A Time-stamp exponentially weighted moving average forecasting method is studied to reduce or eliminate the influence of the network transmission to control computing. We used control time domain to predict the future control actions. In order to maintain good performance and stability of a closed-loop NCSs, through an innovative method of information buffering in the controller and actuator. The buffering mechanism can also deal with the problems of an event of vacant sampling, data losses or data packet disorder. In order to reduce the network load, an threshold is predetermined in controller's cache. If the substraction between the new predictive values and the old predictive values on the time point k is less than the threshold, the new control sequence is not sent to the actuator, and actuators continue to use the old control sequence.Stability is achieved in the sense of Lyapunov's second method by imposing the standard quadratic function which is transformed from the state regulator, While, the standard quadratic function is an effectively Lyapunov function. Feasibility, which in turn implies stability in NICSs with the suboptimal control strategy.As a new control system, the theoretical studies of NCSs involve many subjects. How to combine the research achievements of various subjects to continuously improve the theoretical system of NCSs and broaden its application field is worthy of considering in the future research.