Research On The Robust Control And Filtering For Networked Control Systems

With the increasing research on networked control systems(NCSs) by domestic and foreign scholars, the obtained achievements lay a theoretical foundation for the development of network control systems. The information transmission time delay problem caused by the network must be considered during the study and the phenomenon on the system states that cannot be measured or difficult to measure is also one of the key problems. At the same time, we need to pay attention to the variety of model parameters and consider the uncertainty case in the control system design. In the research process of NCSs, the study is gradually in-depth developments, which varies from the normal controlled object to the complex(such as the generalized object and the linear parameter varying object, etc.), from the continuous/discrete time to the hybrid time, from linear systems to nonlinear systems. In addition, the corresponding control theory also unceasingly develops from state feedback control to the observer-based state feedback control and from the controller design to the filter design. Taking this train of thought as the main line, this paper has carried on the related research on NCSs, and the proposed methods have been proved to be effective through the numerical examples. The main contents are as follows:First of all, in view of the linear NCSs with different controlled objects, the observer-based robust controller design problem is investigated. The observer-based H? control for normal linear NCSs, the sampled-data control based on the observer for linear hybrid NCSs and the observer-based H? and 2 ?L? L control methods for generalized NCSs are proposed, respectively.Secondly, a new linear parameter varying network control structure is considered. Under the network environment, the model analysis is made by combining with the characteristics of LPV(Linear Parameter Varying) system, and robust H? controller design method is given with the state feedback control strategy. Then, the case that system states are not fully measurable is studied. The model is augmented by introducing the state observer and the problem of observer-based H? controller design is converted into the controller and observer collaborative design problem with two independent time-varying delays.Next, the robust filtering problem for nonlinear NCSs described with T-S fuzzy model is investigated. Through the filtering strategy analysis of stochastic time-delay systems based on T-S fuzzy approximation, the Hankel norm criterion is first established which can guarantee the corresponding filtering error system to be mean-square asymptotically stable and to satisfy the specified performance level. And then the discrete and continuous situations for nonlinear NCSs are discussed, the robust H? filtering method of discrete-time NCSs and the robust 1L filtering method of continuous-time NCSs are presented, respectively.Finally, the norm-bounded uncertainties are considered for the nonlinear NCSs. Based on the T-S fuzzy model, the solutions to the robust H? filtering problem of discrete uncertain NCSs and the robust 1L filtering problem of continuous uncertain NCSs are given, respectively.