基于LM-Gaussian算法的雷达测速方法研究  

Research on Radar Speed Measurement Method Based on LM-Gaussian Algorithm

作　　者：孙杰 SUN Jie(Hangzhou Brolight Technology Co.,Ltd.,Hangzhou,Zhejang 310000,China)

机构地区：[1]杭州博源光电科技有限公司,浙江杭州310000

出　　处：《自动化应用》2025年第6期216-218,共3页Automation Application

摘　　要：在民用场景下,毫米波雷达测速的关键是将雷达天线模块输出的中频信号提取后,通过快速稳定的算法处理,得到实时准确的多普勒频移信号,继而计算出目标物的运动速度。提出了一种将高斯拟合嵌套在LM算法中,通过采样数据计算自动设置合适的初始条件,并进行非线性最小二乘拟合快速迭代运算,最终收敛并输出多普勒频移参数的方法。实验结果表明,即使在相对波动的条件下,该方法仍能获得稳定精确的测量速度。In civilian scenarios,the key to millimeter wave radar speed measurement is to extract the intermediate frequency signal output by the radar antenna module,process it through a fast and stable algorithm,obtain real-time and accurate Doppler frequency shift signals,and then calculate the motion speed of the target object.This paper proposes a method of embedding Gaussian fitting into the Levenberg-Marquardt(LM)algorithm,automatically setting appropriate initial conditions through sampling data calculation,and performing fast iterative operations of nonlinear least squares fitting,ultimately converging and outputting Doppler shift parameters.The experimental results show that even under relatively fluctuating conditions,this method can still achieve stable and accurate measurement speeds.

关 键 词：雷达测速 LM算法 高斯拟合 多普勒频移 

分 类 号：TN98[电子电信—信息与通信工程]