Research On Detection Method Of Sleep Vital Signs Based On Fan-shaped Distributed Radar

Breathing and heartbeat are two important vital signs of the human body,which can reflect the health of the human body.The sleep process accounts for about one-third of the average person’s daily time,and it is of great significance to detect the vital signs of the sleep process.At present,most of the devices applied to the detection of sleep vital signs are contact devices,and the contact devices have a lot of interference and inconvenience to the human body.In recent years,non-contact vital signs detection methods have been paid more attention,especially the radar-based vital signs detection technology,which has achieved some research results,but still has the problem of insufficient detection accuracy,and cannot handle the complicated sleep postures interference to detection.Aiming at the above problems,this thesis proposes a sleep vital sign detection method based on fan-shaped distributed radar.The main work and contributions are as follows:(1)Aiming at the problem of insufficient accuracy of vital signs detection with single radar,this thesis proposes a scheme to detect human vital signs using multiple radars,and designs and builds a fan-shaped distributed radar system.The fan-shaped distributed radar system includes radar front-end modules,filter amplifying modules,and acquisition and transmission module.(2)In terms of respiration rate and heart rate extraction,this thesis firstly uses arctangent demodulation algorithm to demodulate the radar output baseband signal,and then uses wavelet transform to separate the respiration signal and heartbeat signal from the demodulated signal.Correlation function method and windowed Fourier transform method to estimate respiratory rate and heart rate,respectively.Aiming at the IQ imbalance problem of the radar output baseband signal,this thesis uses the ellipse fitting method to calibrate the imbalance of the baseband signal before performing arctangent demodulation.Aiming at the insufficient frequency resolution problem for short-term heartbeat signal frequency estimation,the frequency domain is interpolated to improve the frequency resolution.Based on the fan-shaped distributed radar system,multiple detection results of respiratory rate and heart rate can be finally obtained for each detection.(3)In the fusion of respiratory rate and heart rate,a fusion method of vital sign frequency estimation based on signal-to-noise ratio analysis is proposed.In this thesis,the multiple respiratory rate or heart rate values detected by the fan-shaped distributed radar radar are screened,and the weights are configured for the screened respiratory rate or heart rate by calculating the signal-to-noise ratio of the respiratory signal or the heartbeat signal,and then fused respiration rate or fused heart rate can be obtained by weighted fusion.Multiple short-term vital signs detection experiments in three sleep postures show that mean square errors of the fusion method proposed in this thesis is 0.19 and 0.15 smaller than the fusion method based on signal amplitude analysis and signal energy analysis in respiratory rate detection,respectively.In heart rate detection,the mean square error of the fusion method proposed in this thesis is 2.61 and 4.31 smaller than the two method,respectively.In the heart rate continuous monitoring experiment,the mean square error of the fusion method proposed in this thesis is 2.02 and 3.23 smaller than that of the fusion method based on signal amplitude analysis and signal energy analysis.