复合高斯杂波中基于少量辅助数据的自适应目标检测算法研究

The radar detection ability is severely affected by the detection environment complexity. As the key method to improve radar target detection performance, adaptive detection technology is able to estimate the clutter covariance matrix in real time and implement the accurate matching between detection statistics and clutter parameters. For the compound-Gaussian clutter background, this project use the a prior information of clutter to solve the problems of clutter covariance matrix estimation and adaptive target detection in the case of limited secondary data. Firstly, from the beginning of clutter statistical property analysis, the a prior model and structure characteristics of clutter covariance matrix are utilized, and the reconstructed-model-based and structure-information-based clutter covariance matrix estimation algorithms are proposed; then, the mathematical model of compound-Gaussian clutter and clutter covariance matrix estimation method are integrated, and the a-prior-information-based target adaptive detection algorithm is proposed; finally, the simulations are designed and clutter measured data are used to validate the algorithm performance. The research in this project can enhance the radar detection ability of weak targets in complex environments, which is of great academic research importance; its theory and technology can be used in the detection of targets, such as sea/ground-skimming missiles, cruise missile, small unmanned aerial vehicles, ships and so on, which have broad use in many areas.

现代雷达探测环境的复杂化严重影响雷达的探测能力。自适应检测技术可以利用辅助数据对杂波协方差矩阵进行实时估计，实现检测统计量与杂波参数的精确匹配，是改善雷达目标探测性能的关键技术。本项目针对复合高斯杂波背景，利用杂波的先验信息解决少量辅助数据情况下的杂波协方差矩阵估计和自适应目标检测问题。首先，从杂波的统计特性分析入手，分别利用杂波协方差矩阵的先验模型和结构特性，提出基于重构模型和基于结构信息的杂波协方差矩阵估计算法；然后，结合复合高斯杂波的数学模型和杂波协方差矩阵的估计方法，提出基于先验信息的自适应目标检测算法；最后，设计仿真实验并利用杂波实测数据，验证算法性能。该项目的研究可以提高雷达对复杂环境中目标的检测能力，具有重要的学术研究意义；其理论成果和技术可用于掠海/掠地飞行导弹、巡航导弹、小型无人机、舰船等微弱目标的探测，具有广阔的应用前景。

本项目针对少量辅助数据情况下的杂波协方差矩阵估计和自适应目标检测问题展开研究，内容涉及基于少量辅助数据的复合高斯杂波协方差矩阵结构估计算法和基于少量辅助数据的一步自适应检测算法和两步自适应检测算法等内容，突破少量辅助数据情况下的杂波协方差矩阵估计问题和凸优化问题最优解的求解问题等关键科学问题，形成一系列创新研究成果，包括基于杂波协方差矩阵重构模型的一步自适应GLRT检测算法和两步自适应GLRT检测算法以及基于中位矩阵估计的两步自适应GLRT检测算法等方法，发表学术论文8篇，其中在IEEE T-GRS等国际学术期刊发表SCI检索学术论文3篇；在IEEE Radar Conference等国际会议发表EI检索学术论文5篇；申请发明专利3项。该项目的研究提高了我国目标检测技术的基础研究水平和自主创新能力。