独立函数元：模型、方法和应用

摘要

In the paper, we present a new method for signal processing in statistical domain. The multiple components obtained from the conventional linear transformation are possibly irrelevant, and usually do not possess the characteristics of statistical independence. Therefore a method of signal decomposition and reconstruction is proposed based on the independent function element. This method not only inherits the advantages of linear transformation, but also has the capability of signal representation in the statistical domain. In this paper, the model, definition and obtaining method of independent function element are discussed, and the applications of heart sounds independent function element in the heart sound signal processing are also analyzed in detail. The validity and practicability of the method are demonstrated through two experiments.%针对常规线性变换获得的多分量成分可能不相干，但通常不满足统计独立的特点，提出了一种基于独立函数元的信号分解和重构的方法。该方法不仅继承了线性变换的诸多优点，而且还有统计域表征的优点。讨论了独立函数元的模型、定义和获取方法，详细分析了心音独立函数元在心音信号处理方面的应用。实验验证了所述方法的有效性和实用性。