基于远场测量的沉底目标谐振散射特性重构方法研究

Detection and classification of bottom-laid targets in shallow water is usually conducted with imaging sonar. To achieve long-range localization and a wide range of target classification, the detection efficiency can be greatly improved using acoustic techniques. High frequency sonar can provide high resolution image of objects laying on the ocean bottom. The high frequency acoustic image may contain false alarms as it pertains only to the shape of the object, it cannot, for example, differentiate a mine from a large stone or an oilcan having roughly the same shape. .Low frequency sonar can be used at a long distance and thus provides a high rate for detection and imaging over a wide area. Within this frequency band, the object size is in the same order as the working wavelength (resonance domain). The target can thus be detected and distinguished from other type of target based on the frequency response of the echo return, referred to as acoustic color. .The scattering signature reconstruction of targets from the measured scattered field is summarized as the inverse problem. We propose a new low-frequency active sonar target imaging and classification method. This proposal intends to (1) estimate the properties of bottom-laid targets (to differentiate from another type of target) and referred to as inverse problem, and (2) include acoustic color image reconstruction in the study. The standing-off distance is rather large and considering detection speed, the echo signal can only be measured with limited number of angles in the far field. The total field inside the target region will be reconstructed in order to ameliorate the reconstruction quality and to obtain a more precise acoustic color image. A nonlinear inverse problem has consequently to be resolved. One also needs to demonstrate effective bottom-laid target localization and classification in shallow water waveguide, where the interaction between signal propagation and the extended target scattering effect has to be considered in multi-layers medium.

海洋沉底目标的探测和识别所采取的主要手段是成像。用声学手段实现远距离定位和大范围分类，可极大提高沉底目标探测效率。高频声呐获取的高分辨率目标图像，主要反映了目标几何散射特性，无法区分形状相似的目标，如水雷、岩石、油罐等，误判率较高。.低频声呐信号衰减相对较低，可工作在较远距离，且目标尺寸与低频探测波长在同一数量级(谐振区域)，可以根据回波的频率-角度响应（称为声音颜色）对沉底目标分类。.由测量声场提取产生这一声场的散射体的散射特征，归结为逆问题求解。本立项提出一种新的低频主动声呐沉底目标成像方法：1）针对目标受低频声信号激励发生的谐振散射特性，利用测量远场数据，通过重构目标区域总声场分布，提高远场预报精度，获取更精细的声学颜色特征图像，降低目标分类难度。2）沉底目标通常位于浅海区域，远场测量回波受到低频声信号传播和目标散射相互作用影响。因此，需要研究分层介质中目标散射特性重构方法。

海洋沉底目标的探测和识别所采取的主要手段是成像。用声学手段实现远距离定位和大范围分类，可极大提高沉底目标探测效率。高频声呐获取的高分辨率目标图像，主要反映了目标几何散射特性，无法区分形状相似的目标，如水雷、岩石、油罐等，误判率较高。..低频声呐信号衰减相对较低，可工作在较远距离，且目标尺寸与低频探测波长在同一数量级，可以根据回波的频率-角度响应（称为声音颜色）对沉底目标分类。由测量声场提取产生这一声场的散射体的散射特征，归结为逆问题求解。本立项实现了一种低频主动声呐沉底目标成像方法：1）沉底目标通常位于浅海区域，远场测量回波受到低频声信号传播和目标散射相互作用影响。通过高分辨定位和基于数据的模深度函数估计，实现了波导环境目标散射回波提取和频率-角度散射特性重构；2）开发了一种基于互质阵的目标高分辨到达角估计方法，有效解决了波导传播效应引起的到达角估计交叉项问题；3）针对波导环境参数估计，提出了一种解卷积和粒子滤波相结合的高分辨水平波数估计方法。..项目共发表科研论文5篇，其中第一标注4篇，授权国家发明专利1项。项目负责人参与撰写专著1部，并于2020年获得了麻省理工科技评论中国区35岁以下科技创新35人荣誉称号（Tr35）。依托项目取得的研究成果，于2021年度获得了国家自然科学基金面上项目资助1项。