Fowler A C, Kember G, Johnson P, Walter S J, Fleming P, Clements M
Mathematical Institute, Oxford University, U.K.
J Theor Biol. 1994 Oct 7;170(3):273-81. doi: 10.1006/jtbi.1994.1188.
We present a method based on dynamical systems theory which can be used to filter time series in a way which is superior to classical Fourier decomposition. This method is applied to three data-sets, taken from respiratory measurements of two children in quiet and REM sleep. Our purpose is to filter the several different oscillatory mechanisms which operate, in order to provide clearer signals on which further analysis and diagnosis can be based.
我们提出了一种基于动力系统理论的方法，该方法可用于以优于经典傅里叶分解的方式对时间序列进行滤波。此方法应用于三个数据集，这些数据集取自两个儿童在安静睡眠和快速眼动睡眠状态下的呼吸测量。我们的目的是对运行的几种不同振荡机制进行滤波，以便提供更清晰的信号，在此基础上进行进一步的分析和诊断。
