基于降维状态观测器的最优拥塞控制研究  

Study of optimal congestion control based on reduced-order state-observer

作　　者：韩存武[1] 刁奇[1] 常舒瑞 毕松[1] 刘蕾[1] 庞中华[1] 

机构地区：[1]北方工业大学现场总线及自动化北京市重点实验室,北京100144

出　　处：《高技术通讯》2015年第12期1047-1052,共6页Chinese High Technology Letters

基　　金：国家自然科学基金(61573024;61203230);北京市自然科学基金(4142014;4154068;4152014)资助项目

摘　　要：在Internet拥塞控制系统建模的基础上,将降维状态观测引入到拥塞控制中,进而提出了一种基于主动队列管理(AQM)的新的最优拥塞控制方法。该方法利用降维状态观测器来获取无法直接获取的系统状态,从而实现状态反馈,克服传统的基于AQM的网络拥塞控制方法无法直接得到状态反馈控制律的缺陷,并能通过求解线性矩阵黎卡提方程,得到使性能指标达到最优的状态反馈控制律,实现Internet的最优拥塞控制。仿真结果表明,这种新的最优拥塞控制方法具有很好的跟踪性能。Based on the modeling of Internet congestion control was introduced into the congestion control, and then a new systems, the concept of reduced-order state-observation optimal congestion control method based on active queue management （AQM） was presented. The new method uses a reduced-order state-observer to obtain the system state which cannot be obtained directly, so as to realize the state-feedback to overcome the drawback of traditional AQM based congestion control methods that they cannot directly obtain the state-feedback control law. And it can obtain the sate-feedback control law capable achieve the optimal congestion control tracking performance. of optimizing the performance index by solving the Riccati equations, and for Internet. The simulation results verified the proposed method＇ s good

关 键 词：拥塞控制 最优控制 黎卡提方程 主动队列管理(AQM) 状态观测器 

分 类 号：TP393.06[自动化与计算机技术—计算机应用技术]