摘要
提出了一种基于G.P算法的关联维计算中分形无标度区的识别方法。理论分析和实际应用表明,该方法可利用计算机编程实现,并能根据实际需要客观、有效、准确地识别出时间序列的分形无标度区,计算效率高,人为因素对计算结果的影响小,特别是对于信号分析中关联维计算频繁或无标度区较窄、用常规算法难以准确识别的场合。
In this paper, a novel method of identification the fractal scaleless band of timeseries is proposed based on the study on the correlative dimension of fractal construction and the theory of reconstructing phase space. The result of theory analysis and application shows that the method can be realized by using a computer and with the method the fractal scaleless band of timeseries can be identified according to the parameters given. When the fractal scaleless band identified is narrow, the method is efficacious, objective and precise in identifying the fractal scaleless band of timeseries.
出处
《弹箭与制导学报》
CSCD
北大核心
2003年第1期35-38,共4页
Journal of Projectiles,Rockets,Missiles and Guidance
基金
国家"九五"
"十五"国防重点预研项目。
关键词
G.P算法
分形无标度区
关联维
时间序列
G.P algorithm
fractal scaleless band
correlative dimension
time-series
