摘要
基于小波能对信号的各个频率段进行分离的原理,设计了基于小波包分析的变点探测算法,并以此来研究时间序列的突变.通过对长江宜昌站的年径流水文时间序列进行实验分析发现,在过去120a,长江宜昌站的年最小流量序列和年平均流量序列的均值都极有可能存在着突变,而且其变化趋势都是均值明显减小.该方法不需要对时间序列作任何概率分布和相依性的假定,便能方便地通过调节小波包变点探测算法的参数应用于其他领域.
Based on the principle that different frequency signal can be separated from each other by wavelet, a change-points detection algorithm on the basis of wavelet packet analysis is designed to study the abrupt change of the mean value of time series. It is applied to detect change-points of the annual discharge series of Yangtze river at Yichang hydrological station. The results show that, during the past 120 years, the mean values of both the annual minimum discharge series and the annual average discharge series are very likely to have changed abruptly, and the change tendency for both series is that the mean value has significantly reduced. This algorithm does not need any probability distribution and interdependent hypothesis to the time series, and can be conveniently applied to other fields by regulating its parameters.
出处
《控制与决策》
EI
CSCD
北大核心
2005年第5期521-524,共4页
Control and Decision
基金
国家重大基础研究前期专项项目(2003CCA00200)
国家重点实验室开放基金项目(2004B009)
湖北省自然科学基金项目(201130485).
关键词
小波包分析
时间序列
交点分析
交点探测算法
Data processing
Hydrologic instruments
Probability distributions
Statistical methods
Time series analysis
Wavelet transforms
