东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 2012, Vol. 33 ›› Issue (10): 1377-1380.DOI: -

• 论著 • 上一篇    下一篇

基于PSO优化LS-SVM的GPRS工业控制网络时延预测

田中大;高宪文;李琨;   

  1. 东北大学信息科学与工程学院;
  • 收稿日期:2013-06-19 修回日期:2013-06-19 出版日期:2012-10-15 发布日期:2013-04-04
  • 通讯作者: -
  • 作者简介:-
  • 基金资助:
    国家自然科学基金资助项目(61034005)

Time-delay prediction based on LS-SVM optimized by PSO for GPRS industry control network

Tian, Zhong-Da (1); Gao, Xian-Wen (1); Li, Kun (1)   

  1. (1) School of Information Science and Engineering, Northeastern University, Shenyang 110819, China
  • Received:2013-06-19 Revised:2013-06-19 Online:2012-10-15 Published:2013-04-04
  • Contact: Tian, Z.-D.
  • About author:-
  • Supported by:
    -

RichHTML

0

PDF (PC)

685

摘要： 针对GPRS工业控制网络,采用Socket通信方式搭建了测试平台,在此平台上使用TCP和UDP两种协议对GPRS网络实际时延进行了测试和分析,给出了现场应用中的指导意见.基于时间序列分析,采用粒子群优化的最小二乘支持向量机的方法对GPRS工业控制网络时延进行了预测.仿真结果表明,该方法能较好地预测GPRS网络的时延,为之后的网络预测控制提供了良好的基础.

关键词: 通用分组无线服务, 粒子群优化, 最小二乘支持向量机, 时延, 预测

Abstract: The test platform was implemented with Socket communication method for GPRS industry control network, and the time-delay of the GPRS industry control network was tested and analyzed by TCP and UDP protocols. The guiding opinion was given in the field application. On the basis of time series theory, the time-delay of GPRS industry control network was predicted by using the LS-SVM optimized by PSO. The simulation results showed that the proposed method could very well predict the time-delay of the GPRS industry control network. It also provided a good foundation for the following network predictive control.

中图分类号: