Computer Engineering and Applications ›› 2013, Vol. 49 ›› Issue (20): 91-95.

Previous Articles     Next Articles

Research of NCS forward time delay prediction based on ARMA model

XU Pei, CHEN Qigong, GE Yuan, JIANG Rongrong   

  1. Anhui Provincial Key Lab of Detection Technology and Energy Saving Devices, Anhui Polytechnic University, Wuhu, Anhui 241000, China
  • Online:2013-10-15 Published:2013-10-30

基于ARMA模型的NCS前向时延预测研究

许  培,陈其工,葛  愿,蒋蓉蓉   

  1. 安徽工程大学 检测技术与节能装置省级重点实验室,安徽 芜湖 241000

Abstract: The inherent network time delay in networked control system will reduce the performance of system and even lead to instability. Network time delay mainly includes forward time delay and backward time delay. Considering that forward time delay has not occurred when controller designing the control law, this paper uses Auto-Regressive and Moving Average model(ARMA) to make forward time delay prediction and compares it with the prediction effect of Radial Basis Function(RBF) neural network, the validity and superiority of the method has been demonstrated.

Key words: networked control system, forward channel time delay, prediction, Auto-Regressive and Moving Average model(ARMA)

摘要: 网络化控制系统(Networked Control System,NCS)中固有的网络时延会降低系统性能甚至导致系统不稳定,网络时延主要包括前向时延和后向时延。考虑到前向时延在控制器设计控制律时尚未发生,采用自回归滑动平均模型对前向时延进行预测,并将其与径向基函数(RBF)神经网络预测结果进行了对比分析,验证了所给方法的有效性和优越性。

关键词: 网络化控制系统, 前向通道时延, 预测, 自回归滑动平均模型(ARMA)