基于改进经验模态分解的雷达生命信号检测

Radar vital signal detection based on improved complete ensemble empirical mode decomposition with adaptive noise

从线性调频连续波(FMCW)雷达中提取的生命信号包含大量的噪声,为了获得高信噪比的呼吸和心跳信号,提出了一种基于改进的自适应集合经验模态分解(ICEEMDAN)的生命信号检测方法。该方法首先对FMCW毫米波雷达获取的生命信号进行ICEEMDAN分解,得到若干个固有模态函数(IMF)分量,然后利用IMF分量滤波器选择频率峰值在呼吸和心跳频带范围内的IMF分量,最后根据与雷达生命信号的相关性从滤波的结果中选择IMF分量重构呼吸和心跳信号。实验结果表明,所提出的方法能够准确地检测到心跳和呼吸信号,提取得到的呼吸和心跳信号具有良好的信噪比。

Vital signal extracted by frequency modulated continuous wave(FMCW) millimeter wave radar contains a lot of noise. In order to obtain respiratory and heartbeat signals with high SNR, an algorithm based on improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN)is proposed. Firstly, ICEEMDAN is applied to decompose the radar vital signal acquired by FMCW millimeter-wave radar, and several intrinsic mode function(IMF) components are obtained. Then, IMF component filters are used to select the IMF components with frequency peaks in the frequency range of respiration and heartbeat. Finally, according to the correlation relationship between IMF components and radar vital signals, the IMF components are selected from the filtering results, and used to reconstruct the respiratory and heartbeat signals. The experiments show that the proposed method can detect heartbeat and respiratory signals accurately, and the respiratory and heartbeat signals have good SNR.