| 基于AR模型的WSN流量多步预测算法研究 |
| |
| 引用本文： | 叶廷东.基于AR模型的WSN流量多步预测算法研究[J].广东轻工职业技术学院学报,2011(4):1-5. |
| |
| 作者姓名： | 叶廷东 |
| |
| 作者单位： | 广东轻工职业技术学院计算机工程系 |
| |
| 基金项目： | 东莞市科技计划项目(20100815400156);广东轻工职业技术学院2010年科技基金(KJ201022) |
| |
| 摘 要： | 针对WSN流量预测,基于AR模型提出一种WSN流量双卡尔曼并行递推预测算法,该算法使用两个Kalman滤波器,交替进行AR模型参数的递推辨识与时变数据中真实值的最优估计,根据序列数据的最新信息实时修正AR模型参数进行动态预测。同时针对大步长的流量预测,引入滚动修正思想,克服动态预测算法存在间隔时间过长的缺点,降低多步预测误差。实验研究表明,利用研究的双卡尔曼并行递推算法使用AR模型进行多步预测,从原理设计和实现算法上,实现了WSN流量的准确预测。
|
| 关 键 词： | AR模型 无线传感器网络 卡尔曼 预测 网络流量 |
Multi-step Prediction Algorithm Research of WSN Network Traffic Based on AR Model |
| |
| Ye Tingdong.Multi-step Prediction Algorithm Research of WSN Network Traffic Based on AR Model[J].Journal of Guangdong Industry Technical College,2011(4):1-5. |
| |
| Authors: | Ye Tingdong |
| |
| Institution: | Ye Tingdong(Compute Engineering Dept.,Guangdong Industry Technical College,Guangzhou,510300) |
| |
| Abstract: | Aimed to the prediction of WSN network traffic,the paper advanced a REPK dynamic prediction algorithm of WSN network traffic based on AR model;the algorithm used two Kalman filter to recursively identify parameters of AR model and optimally estimates actual data in time-varying data and it can use fresh information to amend parameters of AR model real-timely and predict dynamically.And aimed to the prediction of time series signal with long delay,the paper applies scroll-amendment method that solves great intervals problem of REPK algorithm and decreases multi-step prediction error.The experiment results show the proposed REPK estimation algorithm based on AR model realizes the accurate prediction of WSN network traffic by principle design of prediction method and skillful algorithm. |
| |
| Keywords: | AR model wireless sensor networks(WSN) kalman prediction network ttraffic |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
