基于VMD的FMCW毫米波雷达胸壁微动检测

VMD⁃based chest wall micromotion detection with FMCW millimeter wave radar

人体的呼吸和心脏跳动导致的胸壁微动可通过毫米波雷达的相位变化进行记录。为了从相位中提取胸壁的微动历史,估计出呼吸和心跳的频率,提出了一种生命体征信号的提取与分离方法。首先,对毫米波雷达中频信号进行傅里叶变换,确定胸壁所处距离单元;接着,针对此距离单元中的相位进行提取与解缠绕,将相位转换为距离变量后得到包含呼吸和心跳信息的生命体征信号;最后,利用变分模态分解(VMD)算法从生命体征信号中分解得到的模态函数中确定呼吸信号和心跳信号,再采用傅里叶变换估计两者的频率。实验中,采集30 s的人体数据进行验证,成功提取到了生命体征信号波形,并分离出了呼吸和心跳信号,呼吸和心跳频率误差在8%以内。分析实验结果表明,文中所提的信号提取与分离方法能得到有效的生命体征信号,并从中估计出较为准确的呼吸和心跳频率。

The chest wall micromotion caused by the respiration and heartbeat of the human body can be recorded by the phase change of the millimeter⁃wave radar.In order to extract the micromotion historical information of the chest wall and estimate the respiration and heartbeat frequencies from the phase,a method of extracting and separating vital sign signals is proposed.Fourier transform is performed for the intermediate frequency signals of the millimeter⁃wave radar to determine the distance unit where the chest wall is located.The phase in this distance unit is extracted and unwrapped,and then is converted into a distance variable to obtain the vital sign signals containing respiration and heartbeat information.Variational mode decomposition(VMD)algorithm is used to discriminate the respiration and heartbeat signals from the modal function decomposed from the vital sign signals,and then the Fourier transform is used to estimate the frequencies of the two.In the experiment,the human body data was collected for 30 s to verify the method,the vital sign signal waveform was extracted successfully,the respiration and heartbeat signals were separated,and the errors of the respiration and heartbeat frequencies were all within 8%.The analysis of the experimental results shows that the signal extraction and separation method proposed can be used to obtain effective vital sign signals,and the respiration and heartbeat frequencies can be estimated from the vital sign signals more accurately.