摘要: |
提出了一种用于混沌时间序列预测的改进型加权一阶局域法。用衰减系数对分维
指数加权一阶局域法的向量距离公式进行修正,调节邻近点与中心点的相关性,也调节了同
一邻近点的各个分量和中心点的最后一个分量的关联程度。利用该方法对Logistic混沌时间
序列进行预测的结果表明,衰减系数取最佳值时,相对于现有算法,该方法可以更精确地
预测混沌时间序列。 |
关键词: 混沌时间序列 预测模型 加权一阶局域法 衰减系数 |
DOI: |
|
基金项目: |
|
An Improved Adding-weight One-rank Local-region Method for Prediction of Chaotic Time Series |
QIAN Feng,WANG Ke-ren,FENG Hui,JIN Hu |
(Electronic Engineering Institute, Hefei 230037, China) |
Abstract: |
This paper proposes an improved adding-weight one-rank local-region method f
or prediction of chaotic time series. An attenuation coefficient i
s applied to amend the vector distance formula of the dimension-exponent adding
-
weight one-rank local-region method. The attenuation coefficient not only adju
st
s different relevance of each adjacent point and the center point, but also adju
sts the correlation between each dimension of the same phase point and the last
dimension of the center point. The Logistic chaotic time series are forecasted u
sing the improved method, and simulation results show that the prediction accura
cy is improved with the optimal attenuation coefficient in the proposed method c
ompared with the original one. |
Key words: chaotic time series prediction model adding-we
ight one-rank local-region method attenuation coefficient |