具有随机传输时滞的网络化控制系统故障检测  

Fault Detection of Networked Control Systems With Stochastic Transfer Delays

作　　者：冒泽慧[1] 周东华[1] 吉吟东[2] 何潇[2] 

机构地区：[1]清华大学信息科学技术学院,北京100084 [2]清华大学信息技术研究院,北京100084

出　　处：《南京航空航天大学学报》2011年第B07期32-35,共4页Journal of Nanjing University of Aeronautics & Astronautics

基　　金：国家自然科学基金(60931160440;61074084)资助项目;中国博士后基金(20100470322;20100470294)资助项目;重庆交通大学教育部重点实验室开放基金(SLK2010A02)资助项目

摘　　要：针对具有干扰的网络化控制系统提出了一种鲁棒故障检测方法。考虑网络化控制系统中,被控对象的传感器和执行器之间的信息传输是通过一个共享传输媒介与控制器连接实现,传感器-控制器,控制器-执行器之间的时滞由两个独立的马氏随机过程分别描述。通过对被控系统状态和故障检测滤波器的适当扩维,鲁棒故障检测滤波器以一组线性矩阵不等式(LMIs)的形式给出设计方法。残差基于控制器和故障检测滤波器获得的数据生成,在无故障发生时,其输出趋于零,故障发生后,跟踪故障信息。最后给出一个例子说明了所提方法的有效性。A robust fault detection （FD） scheme is developed for networked control systems with discurbances. Considering that the sensors and actuators of the plant exchange information with a remote controller via a shared communication medium, two commonly used Markov chains are employed to characterize the delays induced by the network between the sensor-controller and controller-actuator, respectively. After properly augmenting the states of plant and the fault detection filter, an H-infinite fault detection filter can be designed in solving a certain set of linear matrix inequalities （LMIs）. The residual is generated based on the data that can be exactly obtained at the controller and fault detection filter side. If there is no fault,the residual is convergent. Otherwise,it tracks the fault. Finally, an example is given to show the efficiency of the proposed method.

关 键 词：故障检测 网络化控制系统 随机传输时滞 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]