您的位置:首页  > 论文页面

探测气温时间序列的复杂性特征分析方法

发表时间:2009-07-15  浏览量:1969  下载量:805
全部作者: 许娜,商朋见,袁广才
作者单位: 北京交通大学理学院;北京交通大学光电子技术研究所
摘 要: 首先给出离散时间序列多重分形消除趋势涨落的分析(multifractal detrended fluctuation analysis, MF-DFA)方法,并用它们研究了气温时间序列。通过分析发现:气温时间序列具有多重分形性的复杂特征。最后,用MF-DFA方法对北京市气温序列附加正弦函数趋势、幂函数趋势、指数函数趋势和对数函数趋势进行研究,发现附加趋势后多重分形性由强到弱依次是:幂函数、正弦函数、对数和指数趋势。
关 键 词: 动力系统;多重分形消除趋势涨落分析;多重分形谱;气温序列;附加趋势
Title: Detection on the complexity characteristic analysis�methods of temperature time series
Author: XU Na, SHANG Pengjian, YUAN Guangcai
Organization: School of Sciences, Beijing Jiaotong University;Institute of Optoelectronics Technology, Beijing Jiaotong University
Abstract: This paper introduces the multifractal detrended fluctuation analysis (MF-DFA) methods of discrete time series and uses them to study the temperature time series. It is demonstrated that the temperature time series has a complexity of multifractal behavior. Finally, the MF-DFA method is used in Beijing’s temperature time series with four-types of trends: sinusoidal, power-law, exponential and logarithm trends. Additional trends of multifractal from strong to weak are: the power-law, sinusoidal, logarithm and exponential trends.�
Key words: dynamical system; multifractal detrended fluctuation analysis; multifractal spectrum; temperature time series; superposition trends
发表期数: 2009年7月第13期
引用格式: 许娜,商朋见,袁广才. 探测气温时间序列的复杂性特征分析方法[J]. 中国科技论文在线精品论文,2009,2(13):1350-1355.
 
1 评论数 0
暂无评论
友情链接