Millimeter-wave Radar Human Vital Signs Detection Methods

Millimeter-wave radar vital signs detection technology surpasses the limitations of traditional contact-based detection technology.By performing phase analysis on the radar echoes reflected by the chest cavity,respiratory and heartbeat signals can be detected.This technology offers the advantages of small size and easy deployment,has broad application prospects in areas such as smart homes and health monitoring.The extraction,separation,and parameter estimation of vital signs are key steps in the detection of vital signs using millimeter-wave radar.This thesis focuses on these steps,and conducts theoretical research and experimental verification.The main contents are as follows:(1)The preprocessing of millimeter-wave radar vital sign echoes is studied,and methods such as mean subtraction,phase difference and smoothing are used to remove static clutter and enhance heartbeat signals,effectively extracting rough respiratory and heartbeat signals.(2)An adaptive variational mode decomposition based on Envelope Entropy(EEAPVMD)algorithm is proposed.Combined with heartbeat range,respiratory harmonic and intrinsic modal function energy screening,Particle Swarm Optimization(PSO)algorithm is used to adaptively search for the modal decomposition parameters with the smallest envelope entropy,achieving effective separation and estimation of respiration and heartbeat.(3)An adaptive multi-channel variational mode decomposition(APMVMD)algorithm is proposed,which constructs a multi-channel joint envelope entropy optimal model,and uses the correlation constraint between multi-channel data in adaptive mode decomposition to improve the estimation accuracy of respiratory and heartbeat vital signs.(4)A multi-person vital sign detection method is proposed,which utilizes the multidimensional differences in distance and angle of multiple person signals to achieve the separation of multiple person.Combined with the single-person respiration and heartbeat separation and estimation algorithm,it achieves effective detection of vital signs for multiple people.The effectiveness of the above research content has been verified through dataset constructed by the project team.The results show that the methods proposed in this thesis can accurately and effectively achieve single-person and multi-person vital sign detection.