基于连续小波变换的水文序列变化特征研究  

Study on Change Characteristics of Hydrological Time Series with Continuous Wavelet Transform

作　　者：王文圣[1] 赵太想[1] 丁晶[1] 

机构地区：[1]四川大学水电学院

出　　处：《四川大学学报（工程科学版）》2004年第4期6-9,共4页Journal of Sichuan University (Engineering Science Edition)

基　　金：国家自然科学基金资助项目(50279023;40271024)

摘　　要：为揭示复杂水文现象中的精细结构,应用小波分析的多分辨率分析和分形理论相结合的途径探讨了水文序列变化的统计特征。将水文序列进行连续小波变换,得到小波变换系数,计算出小波变换系数的动力学特征指标———分维随尺度的变化过程。根据分维随尺度的变化特性,探讨了水文序列的奇异性(突变、跳跃等)变化。通过仿真试验和实际径流资料计算分析,表明提出的方法是有意义的,并且可行。In order to describe the detail structure of the complicated hydrological system, the statistic characteristics of hydrological time series have been probed based on the multi-resolution of wavelet analysis and fractal theories. Firstly, the continuous wavelet coefficients of hydrological time series in different scale are acquired with Mexican Hat successive wavelet transform. Secondly, the fractal dimension of the wavelet coefficient in special scale is calculated with the box method. The curve between the fractal dimension of the wavelet coefficient and the scale can be plotted. Finally, the singularity (break, jump, trend) of hydrological time series can be diagnosed from the curve. The simulated data and real runoff time series have been studied with suggested method. The results have shown that the research is meaningful and feasible.

关 键 词：水文序列 小波变换 分维 奇异性 

分 类 号：P333[天文地球—水文科学]