【作者】 刘明；

【导师】 张庆灵；

【作者基本信息】 东北大学 ， 运筹学与控制论， 2006， 硕士

【摘要】 本文研究了当被控对象为线性系统时，网络控制系统(NCS)的稳定性分析与控制器设计问题。主要内容如下： 研究了短时延网络控制系统的保性能控制器设计问题。针对闭环网络控制系统的离散化模型，把回路时延的不确定性转化为被控对象系统方程系数矩阵的不确定性，利用所提出的基于LMI的保性能鲁棒控制器设计方法，设计出一个状态反馈保性能控制器，该控制器可以使得闭环网络控制系统达到鲁棒渐近稳定。 研究了传感器到控制器，控制器到执行器之间同时存在网络，网络诱导总时延小于一个采样周期并存在数据包丢失的网络控制系统的建模与控制器设计问题。此种情况下的网络控制系统可等效为一类由结构事件率约束的异步动态系统；利用Lyapunov函数和线性矩阵不等式方法，给出了网络控制系统指数稳定的充分条件以及动态输出反馈控制器的设计方法，并利用遗传算法求出了NCS事件衰减率的极大值。 研究了传感器、执行器为时钟驱动，控制器为事件驱动，网络诱导时延的上界大于一个采样周期T的NCS的Hoo状态反馈控制器设计的问题。将NCS视为一个时滞系统，构造出一个Lyapunov—Krasovskii泛函。与传统方法不同，在处理V的导数时，不进行放大估计，而是通过引入一些恰当的零项，构造出新形式的LMI，从而获得NCS时滞相关稳定的充分条件。与参考文献[24]的方法相比减弱了保守性。在此基础上，给出了系统风控制器的设计方法。 在同时考虑时延和数据包丢失的情况下，给出了网络控制系统保性能控制器的设计方法。首先介绍了参考文献[52]中的NCS模型，其给出的模型是连续域内的，并且无需对系统进行离散化。当被控对象含有不确定性时，该模型仍然适用。然后利用文献[52]中的NCS模型，设计出了使得系统达到鲁棒指数稳定的状态反馈保性能控制器。更多还原

【Abstract】 In this paper, we study the stability analysis and controller design for the networked control systems, the main content is as follows:The guaranteed cost controller design for NCS with network-induce delay less than one sampling period is given. Aiming at the discrete model of closed-loop NCS, the uncertainty caused by loop time-delay is transformed into the matrix parameter uncertainty of the equation of the controlled plant, according to the guaranteed cost controller design method based on LMI, the state feedback guaranteed cost controller is ultimately designed, which can ensure the closed-loop NCS asymptotical stability.When network is inserted from sensor to the controller and from the controller to the actuator, we consider the modeling and controller design for the NCS with both the whole network-induce delay less than one sampling period and existence of data packet dropout. In fact, such kind of NCS can be modeled as a class of asynchronous dynamical systems constrained by the configuration event rate. Based on Lyapunov function and linear matrix inequation, the sufficient condition for the NCS which is subjected to exponential stability and the design of the dynamic output feedback controller are given. At the end of this section we obtain the maximum of the decay rate using genetic algorithm.When sensor and actuator are time-driven, controller is event-driven, we consider the state feedback controller design for the NCS with the maximum of the whole network-induce delay longer than one sampling period. The NCS is regarded as a time-delay system, and a Lyapunov—Krasovskii function is designed for the NCS. By adding some appropriate zero terms to the deviation of V and constructing a linear matrix inequality, the delay-dependent sufficient condition is derived to ensure the closed-loop NCS’s asymptotical stability. Compared with the precious work in reference [24], our method is less conservative. Further the design method of the H_∞ controller for the NCS is presented .With both the network-induce delay and data packet dropout considered, the更多还原