论文部分内容阅读
在对心率变异性（ＨＲＶ）信号及类似的信号进行分析时往往希望把其中具有分形性质的１／ｆ成份和非１／ｆ成份分解开分别处理。然而，由于１／ｆ成份和非１／ｆ成份在频域上相互重叠，不容易简单地用频域滤波的方法来分解。因此有人根据１／ｆ成份的特点提出了所谓粗粒化谱分析的频域分解方法。由于该方法存在一些缺点，本文根据１／ｆ成份的特点提出了基于小波变换的ＨＲＶ信号的时域分解算法。对仿真的ＨＲＶ信号和实际ＨＲＶ信号的分解的结果证实本文提出的方法优于粗粒化谱分析法。
In the analysis of heart rate variability (HRV) signals and the like, it is often desirable to separate the 1/f component and the non 1/f component having the fractal nature therein. However, since the 1/f component and the non 1/f component overlap each other in the frequency domain, it is not easy to decompose using the frequency domain filtering method. Therefore, according to the characteristics of the 1/f component, a frequency domain decomposition method called coarse-grained spectral analysis has been proposed. Due to the disadvantages of this method, this paper proposes a time-domain decomposition algorithm of HRV signal based on wavelet transform according to the characteristics of 1/f component. The results of the decomposition of the simulated HRV signal and the actual HRV signal confirm that the proposed method is superior to the coarse grained spectral analysis method.