基于模型参考自适应的网络控制系统随机时延补偿  

Stochastic Delay Compensation of Networked Control Systems Based on Model Reference Adaptive Control Method

作　　者：李健勇[1] 赵峰[2] 丁青青[3] 

机构地区：[1]郑州轻工业学院计算机与通信工程学院,河南郑州450001 [2]郑州大学电气工程学院,河南郑州450001 [3]清华大学电机与应用电子技术系,北京100084

出　　处：《郑州大学学报（工学版）》2015年第4期6-9,共4页Journal of Zhengzhou University（Engineering Science）

基　　金：国家自然科学基金资助项目(61473265;60974005);教育部博士点专项科研基金项目(20094101120008);河南省教育厅科学技术重点研究项目(13A520379)

摘　　要：随机时延和数据丢包造成网络控制系统性能下降甚至使得系统不稳定.为了提高网络控制系统的动态性能,提出一种基于模型参考自适应控制的双端Smith预估随机时延补偿控制策略.首先,通过在控制器端和执行器端分别引入Smith预估器,将随机时延补偿控制问题转换为模型匹配问题.然后,设计模型参考自适应控制器使Smith预估模型和真实模型相匹配,实现对随机时延的完全补偿.仿真结果表明,提出的控制策略能够对网络时延进行有效补偿,且具有良好的实时性.Stochastic delay and packet loss decline the performance of networked systems and even make it unstable. This paper proposed a novel stochastic delay compensation strategy based on model reference adaptive control and dual Smith predictor. First,by introducing a joined Smith predictor at the controller and actuator terminal end,the stochastic delay compensation problem is transformed to the model matching problem. Then,a model reference adaptive controlled is designed to make the Smith predictor model and the real system model match with each other and to achieve the full compensation of random delay. The simulation results show that the proposed strategy can effectively compensate the network delay and has better real-time property.

关 键 词：网络控制系统 模型参考自适应控制 随机时延补偿 SMITH预估器 

分 类 号：TQ203.2[化学工程—有机化工]