计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (8): 16-19.

• 博士论坛 • 上一篇    下一篇

基于“松散型”小波神经网络的时间序列预测研究

罗 航,王厚军,龙 兵   

  1. 电子科技大学 自动化工程学院,成都 610054
  • 收稿日期:2007-11-12 修回日期:2007-12-17 出版日期:2008-03-11 发布日期:2008-03-11
  • 通讯作者: 罗 航

Research about time-serial prediction based on relaxing wavelet neural-net

LUO Hang,WANG Hou-jun,LONG Bing   

  1. Automation Engineering College of UESTC,Chengdu 610054,China
  • Received:2007-11-12 Revised:2007-12-17 Online:2008-03-11 Published:2008-03-11
  • Contact: LUO Hang

摘要: 分析了一种新的时间序列预测模型——“松散型”小波神经网络预测模型。在用神经网络分析时间序列预测方法的基础上,用方差分析的统计方法确定样本序列的长度,从而有效地确定神经网络输入层节点数。运用该模型对太阳黑子年平均序列进行小波分解、重构、预测和合成,得到了序列总的预测效果。同时,将新模型与传统BP神经网络模型的预测效果进行了比较,分析了两者出现差异的本质原因。整体反映了将复杂问题简单化处理、将小波多分辨分析同神经网络的非线性逼近功能相结合的思想。这种思想及方法发挥了小波变换和神经网络的各自优势,明显提高了预测精度。

Abstract: In this paper,a new time-serial prediction model called relaxing wavelet neural-net is analyzed.At the same time,it decides the length of sample serial by using a kind of statistic method called variance analysis so that the knot number of input layer can be efficiently decided,which is based on analyzing time-serial prediction by means of neural-net.Using this model,it operates the annual average sunspot serial through wavelet-decomposing,reconstructing,prediction and composing.Then,the total prediction effect is gotten.At the same time,the prediction effect about new model and traditional BP-net is compared and the substaintial cause of prediction difference of both types is analyzed.In total,this paper reflects a kind of thought which simplifies complex problem and combines wavelet’s multi-discrimination with neural-net’s non-linear approaching function.This idea exerted wavelet-transformation and neural-net’s advantage respectively.Thus,the prediction precision is obviously enhanced.