摘要
针对一类具有不确定时变时延和数据包丢失的网络控制系统,提出了一种控制器设计方法。在控制器节点利用状态预测器,当传感器数据丢失时,用预测状态代替当前对象状态;当传感器数据传输成功时对预测状态进行修正,并应用预测状态来构造控制律。在执行器节点采用缓冲技术将控制器-执行器时延转化为固定实验。与状态预测器相结合,将有数据包丢失的长时延网络控制系统建模为有事件率约束的异步动态系统。根据此模型,以矩阵不等式的形式给出了系统指数稳定的充分条件,通过求解矩阵不等式的可行解获取控制律。仿真算例说明了该方法的有效性。
The controller design techniques for Networked Control Systems (NCSs) with uncertain, time-varying network induced delay and data packet dropout was proposed. A state predictor was used in the control node to predict current plant states when sensor data was dropout; when the sensor data transmitted successfully, the predicted states were modified according to the received data, and the control law was constructed using the predicted states. Combined with the state predictor, the NCS with delay which may be longer than one sampling interval and data packet dropout, was modeled as asynchronous dynamical system with event rate constraints. Based on the model, sufficient conditions of exponential stability for NCS were given in the form of matrix inequalities. Through the feasible solution of the matrix inequalities the control law can be obtained. A numerical example demonstrates the effectiveness of the proposed method.
出处
《科学技术与工程》
北大核心
2014年第18期76-80,共5页
Science Technology and Engineering
关键词
网络控制系统
矩阵不等式
数据包丢失
异步动态系统
指数稳定性
networked control systemsnamical system exponential stabilitymatrix inequalitiesdata packet dropoutasynchronous dynamieal system exponential stability