【作者】 桑燕芳； 王栋； 吴吉春；

【Author】 Sang Yanfang;Wang Dong;Wu Jichun;Department of Hydrosciences,School of Earth Sciences and Engineering,Nanjing University;

【机构】 南京大学地球科学与工程学院；

【摘要】 应用常用7个小波系中的共53个小波函数,对高斯白噪声的小波特性进行了揭示和描述。分析结果表明,不同小波函数的白噪声小波分析结果总体上可分为三种情况:小波系数幅值随时间尺度水平(TSL)增大时近似相等、递减和递增。然后通过Monte-Carlo模拟统计试验,对各TSL上小波系数平均值、绝对值最大值和极值绝对值平均值的概率分布特性和变化规律进行了分析和定量描述。由于极值绝对值平均值是白噪声小波系数在不同TSL上的整体变化规律的反映,因此在时间序列小波阈值消噪时应将其作为依据,可提高水文序列消噪结果的合理性和可靠性。更多还原

【Abstract】 The 53 wavelet functions in 7 wavelet series are considered,and the wavelet characters of white noises have been analyzed elaborately.Analyses results indicate that when using different wavelets to analyze noises series,the variation rules of wavelet coefficients along with temporal-scale levels(TSLs) are obviously different,which can be divided into three types:being equal,increasing and decreasing.Then the Monte-Carlo(MC) method is applied to identify and quantitatively describe the probability characters and variation rules of average value,absolute maximum value and average absolute extreme value of wavelet coefficients along with TSLs.Finally,by discussing the MC experiments results in detail,several understandings and suggestions about wavelet de-noising of hydrologic series data have been gained,by which the de-noising results hydrologics series data can become more accurate and reasonable.更多还原