The detection of non-contact human vital signs(such as respiration and heartbeat)based on radar,which has significant importance across various fields.One of these challenges,which the work seeks to solve,is the detection of multiple human targets and the filtering of interference,as well as the elimination of heartbeat signal interference in complex scenarios.To meet this need,a multi-target vital sign detection method,which relies on frequency modulated continuous wave(FMCW)radar,is proposed.After first introducing the radar detection principle and signal model,the entire radar signal processing flow is elucidated,which consists of multi-target distance unit detection,phase extraction,and vital sign signal extraction.In the process of multi-target distance detection,where challenges often arise,an improved Cell Averaging Constant False Alarm Rate(CA-CFAR)algorithm is proposed,a method that not only features adaptive detection thresholds but also solves the problem of repeated distance gate detection near targets in traditional CA-CFAR algorithms,thus achieving multi-target detection.During the vital sign signal extraction process,the Automatic Multi-Scale Peak Detection(AMPD)algorithm is utilized so that breathing and heart rates can be estimated,filtering out invalid peak interferences,which enhances the robustness of heartbeat and respiration rate detection.Finally,through multiple sets of tests,which are conducted to verify the methodology,the average error rates for breathing rate and heart rate detection are found to have reduced by 3.67％and 3.31％,respectively,when compared to commonly used spectral analysis estimation methods,the feasibility and accuracy of the proposed method are validated.
维普
万方数据
