摘要
提出并研究基于Internet的远程控制仿真,在测量网络时延实际数据的基础上,利用滑动模型和样条原理,采用滑模均值代替滑模的常规输入值,利用径向基神经网络建立了网络时延的模型.仿真结果表明:该模型能够反映并预测该测量数据所代表的网络路径之间的时延特性,并能在基于Internet的闭环控制系统的设计中有效地替代实际网络进行研究.该建模方法具有快速、准确的特点,能用于在线学习网络模型并对网络的时延值进行预测,为基于Internet的远程闭环控制提供了新的思路.
Realtime telecontrol of mechanical dynamic information with the Internet is attempted.The author analyses the ingredient causingthe time-delay,measures the time-delay practically,uses the Slide-mode and spline func-tion principles for reference,and substitutes the normal Slide-mode inputswith the mean of the Slide-mode,then adoptsthe RBF(radial basis functions) neural network setting up network time-delay model.The simulation results show thatthe model can effectively substitute the actual network in Internet-based closed loop control system research under time-delay condition;the methods used in the modeling,due tothe virtue of speediness and exactness,is capable of learn andpredict the network time-delay on line,this gives a new idea in Internleet-based remote closed-loop control.
出处
《微电子学与计算机》
CSCD
北大核心
2011年第7期41-45,共5页
Microelectronics & Computer