基于凸优化的反卷积相干目标识别算法  

Deconvolutional coherent object identification algorithm based on convex optimization

作　　者：杨东升[1] 郭晓彤[1] 宋建梅 吴桐 YANG Dongsheng;GUO Xiaotong;SONG Jianmei;WU Tong(China Research Institute of Radiowave Propagation,Qingdao 266107,China;Unit 63893 of PLA,Luoyang 471000,China)

机构地区：[1]中国电波传播研究所,山东青岛266107 [2]中国人民解放军63893部队,河南洛阳471000

出　　处：《现代电子技术》2024年第17期73-78,共6页Modern Electronics Technique

摘　　要：为提高反卷积声源成像算法目标识别性能,改善其对相干目标识别效果不理想的问题,提出基于凸优化的反卷积相干声源识别算法。该算法利用阵列接收信号与导向向量共轭转置相乘得到波束输出,去除互谱操作,避免忽略互谱矩阵交叉项中的相干声源信息;其次建立波束输出、目标分布与点扩散函数线性方程组,利用凸优化方法实现目标源强的高精度求解。通过模拟仿真和实验表明:提出的基于凸优化的反卷积相干目标识别算法能有效提高反卷积目标成像算法空间的分辨率和动态范围,在识别相干目标方面更具优势。A convex optimization based deconvolution approach for the mapping of coherent acoustic source(CVX-DAMAS-C)is proposed in this paper to improve the object identification performance of the deconvolutional acoustic source mapping algorithm and perfect its unsatisfied effect of coherent object identification.In the proposed algorithm,the beam output is obtained by multiplying the received signals of the array and the conjugate transpose of the guide vector,the cross-spectral operation is removed,and the coherent acoustic source information in the cross-term of the cross-spectral matrix is avoided.And then,the linear equation sets of beam output,object distribution and point spread function(PSF)are established.The convex optimization method is used to achieve high-precision solution of the object source intensity.The simulations and experiments show that the proposed CVX-DAMAS-C algorithm can improve the spatial resolution effectively and enlarge the dynamic range(DR)of the deconvolutional object mapping algorithm,so it has more advantages in identifying coherent objects.

关 键 词：反卷积 声源识别 目标识别 互谱矩阵 凸优化 点扩散函数 

分 类 号：TN911-34[电子电信—通信与信息系统]