基于形状和位置可调非参数原子的信号波形提取方法

提出一种基于非参数原子的信号波形提取方法．将具有先验知识的模板信号通过一个均匀滤波器组得到一组基函数，并由此构造出用离散序列表示的非参数原子．滤波器组的采用以及子带信号的延时扩展使得构造的波形原子的形状和位置可以自适应地调节．基于奇异值分解技术导出了用非参数原子实现信号波形提取的算法．所提出的方法可以构造出无法用公式和参数描述的波形原子，可从信号中提取任意复杂的波形．实验结果表明，即使在模板信号的幅度、相位或频率与待分析波形存在较大差异时，仍可有效地将相应波形从信号中提取出来．

Signal Waveform Extraction Based on Shape- and Position-Adapted Nonparametric Waveform Atom

A method for signal waveform extraction is proposed, which is based on nonparametric waveform atom. Using a template signal that contains a prior information, a set of basis functions is obtained firstly by means of a uniform filter bank and then a nonparametric waveform atom that is represented by a set of discrete sequences is constructed. The filter bank makes the waveform atom shape-adapted, and a delay expansion of subbands makes the waveform atom position-adapted. Using the atom, an algorithm for extracting waveforms from signal is presented based on singular value decomposition. The proposed method can be used for constructing a waveform atom that can not be characterized by expressions and parameters, and thus for extracting any complicated waveform from signal. Examples show that the waveform to be analyzed can effectively be extracted even if there is a visible discrepancy between the amplitude, phase or frequency of the template signal and those of the waveform.