摘要
不确定性的网络时延导致了远程控制品质降低,成为影响网络化控制系统性能的关键因素。因此,对网络时延的特性分析以及精确的预测对改善远程控制的动态性能起到重要的作用。在对互联网时延进行了大量的测试的基础上,分析了网络时延的原因以及规律,提出了一种基于误差修正的改进BP神经网络来预测互联网不确定性时延的方法。通过实验比较,证明了该预测方法的有效性。
The uncertain time delay of the network, which caused worse quality of remote control, becomes the main factor that affects the performance of networked control systems. Thus, analyzing characteristics of time delay of the network and accurately predicting the time delay play key role in improving dynamic performance of remote control. On the basis of plenty of tests on time delay of Interact, the reasons and rules of time delay of network are analyzed. The method of using modified BP neural network based on error correction is proposed to predict uncertain time delay of Interuet. The results of experiment prove the effectiveness of prediction.
出处
《自动化仪表》
CAS
2008年第3期1-4,共4页
Process Automation Instrumentation
关键词
网络化控制系统
变时延
时延预测
改进BP算法
精度
Networked control system Variable time delay Prediction of time delay Modified BP algorithm Accuracy
