Research And Application Of Multi-target Detection Algorithm For Millimeter Wave Radar

Millimeter-wave radar has broad prospects and unique applications in the field of target detection sensors,and its excellent performance,including high resolution,high accuracy,strong anti-interference capability and privacy protection,has attracted much attention.Target detection algorithms identify and localize targets sensed by millimeter-wave radar,enabling the system to accurately track and analyze the target.This is critical for many application scenarios,including smart driving,security surveillance,and UAV navigation.The performance of millimeter-wave radar systems can be continuously improved and optimized to enhance the accuracy and reliability of target detection through an in-depth study of target detection algorithms.In this paper,based on frequency modulated linear continuous wave(LFMCW)millimeter wave radar,we investigate the optimization and improvement of millimeter wave radar digital signal processing algorithms and propose algorithm designs for two specific application scenarios:multi-target tracking and vital sign detection.The main tasks are as follows:(1)The LFMCW radar signal preprocessing algorithm is investigated.In order to overcome static object noise and dynamic object noise in the surrounding environment,a clutter suppression algorithm with moving target display(MTI)and moving target detection(MTD)cascade is proposed to effectively suppress static and dynamic clutter.To maintain a constant false alarm probability,an FOSCA-CFAR constant false alarm detector is proposed to avoid the target masking effect in a multi-target environment and reduce the system computation to achieve stable detection of multiple targets.(2)A multi-target tracking algorithm based on LFMCW millimeter wave radar is proposed.The Ellipse-DBSCAN algorithm is proposed for cluster clustering based on the characteristics of millimeter wave radar point clouds,which reduces the mislabeled discrete noise and proposes an extended Kalman filter-based cluster tracking algorithm for trajectory tracking and state prediction.Finally,the feasibility of this multi-target tracking algorithm is verified by field experiments.(3)The LFMCW millimeter wave radar-based algorithm for vital sign detection is proposed.The small displacement of human chest cavity during respiration and heartbeat is transformed into the change of echo signal phase to estimate the respiration and heartbeat rate.Firstly,a phase unwinding algorithm is proposed for the phase collapse phenomenon existing in the IQ signal solving phase,and design filter as a band-pass filter to separate the respiration and heartbeat signals.Then,the appropriate window length is determined by simulation,and the short-time Fourier transform is used to perform time-frequency analysis on the respiration and heartbeat signals to obtain the real-time respiration rate and heartbeat rate.Finally,the feasibility of this vital sign detection algorithm is verified by comparing the results with the smartwatch detection results.