摘要
基于最大重叠离散小波谱的定义,本文提出了一种确定分形信号局部奇异性指数的算法,构造了一种类似于奇异性谱的直方图,并用之来描述信号奇异性的全局统计分布.算法的有效性通过数字试验及在真实心率数据中的应用得到了验证.
Based on the definition of wavelet spectrum of the maximal overlap discrete wavelet transform, we propose a novel algorithm for determining the local singularity exponents, which provides the local scaling information of interest,and construct a singularity-spectrum-like histogram to describe the global statistical distn＇bution of the local singularity strengths. The validity and efficiency are verified by numerical experiment and application to real heart rate data.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2008年第3期482-485,共4页
Acta Electronica Sinica
