一种基于高斯混合模型抑制声波雷达地物杂波的算法  

A Ground Clutter Suppression Algorithm for Sodar Based on Gaussian Model

作　　者：王强[1] 张皓[1] 冯策[1] 朱明 李悦 WANG Qiang;ZHANG Hao;FENG Ce;ZHU Ming;LI Yue(No.3 Research Institute of China Electronic Technology Group Corporation,Beijing 100015,China)

机构地区：[1]中国电子科技集团公司第三研究所,北京100015

出　　处：《电声技术》2023年第4期137-141,共5页Audio Engineering

摘　　要：声波雷达设备在复杂地形条件测风时,返回信号是大气湍流散射信号和周围环境地杂波的叠加,谱数据中通常会出现明显的双峰结构,且满足高斯模型。通过混合高斯分峰技术区分大气湍流功率谱和地杂波功率谱,结合传统的地物杂波抑制方法降低探测误差。试验研究表明:基于最大期望(Expectation Maximization,EM)混合高斯分峰方法可有效区分地物杂波谱并获取谱特征信息,且进行地杂波抑制后,数据质量得到明显改善。Using the sodar to detect wind in complex terrain,the return information is a superimposition of turbulent signal and ground clutter,the power spectral data would appear an obvious bimodal structure,and can be seen as a double Gaussian form.This paper uses the method of Gaussian ftting to distinguish the power spectrum of atmospheric turbulence signal and the power spectrum of ground clutter,combined with the traditional method of ground clutter suppression to reduce the detection error.The study shows that the Gaussian fitting can distinguish the ground clutter spectrum and obtain the spectral characteristic information effectively.After ground clutter suppression,the data quality was obviously improved.

关 键 词：声波雷达 地杂波 混合高斯分峰 

分 类 号：TN957.51[电子电信—信号与信息处理]