基于频率差分的水下分布式阵列稳健方位估计方法研究

Distributed array which composed of several distributed subarrays can effectively increase the aperture of array and improve the accuracy of direction-of-arrival (DOA) estimation and has great application prospects in underwater target detection. It can be widely used in bottom mounted sonar systems, and is expected to become an important part of the seabed observation network. However, large subarray spacing may lead to serious grating lobes. The structure of distributed arrays is easy to suffer from severe deformation due to the marine environment.. Focusing on the problems of grating lobe ambiguity and mismatch of array errors in underwater distributed array estimation, this project uses frequency difference and self-calibration technology to realize grating lobe suppression and joint calibration of element position and amplitude-phase response errors, respectively. Firstly, a new idea of the grating lobes suppression based on the frequency difference method is proposed. The frequency of the received wideband underwater acoustic signal is reduced with the frequency difference method and the grating lobes are hence suppressed. Secondly, the self-calibration technology is studied, and the DOA estimation method is formed when the array position mismatch and sensor response error coexist. Therefore, the robustness of DOA estimation can be improved even for multiple array errors coexist. Experiments are carried out to verify the performance of the robust DOA estimation method for distributed arrays. The existing underwater data on the sea is also processed to evaluate the proposed methods.

水下分布式阵列由多个分布式子阵组成，能够有效地实现阵列孔径扩展，提高目标方位估计精度，可广泛应用于海底固定式声呐系统，有望成为海底观测网络的重要组成部分。然而，较大的子阵间距会导致严重的栅瓣模糊，同时海洋环境易造成阵列结构严重失配，影响阵列处理性能。. 本项目针对水下分布式阵列的栅瓣模糊以及阵列参数失配等问题，分别利用频率差分和自校准技术，实现栅瓣抑制以及阵元位置与响应误差的联合校准，完成分布式水听器阵列的稳健高分辨方位估计方法研究。首先，针对宽带水声接收信号提出了一种基于频率差分技术进行栅瓣解模糊处理的新思路，利用频率差分方法进行降频处理，抑制栅瓣模糊；其次，深入研究阵列自校准方位估计技术，完成阵元位置严重失配与响应误差并存时的方位估计方法研究，提高多种阵列误差并存时目标方位估计的稳健性。开展实验研究，并结合已有数据资源进行处理与分析，验证研究方法的有效性。

项目旨在研究分布式水听器阵列的稳健方位估计技术，主要解决分布式阵列栅瓣模糊问题以及海洋环境中分布式阵列阵元位置误差、水听器幅相误差等引起的稳健性问题。.经过三年执行周期的研究，本项目阐释了分布式阵列栅瓣产生机理，并给出了基于频率差分抗混叠的处理思路；分析了阵列严重弯曲变形时对阵列流形相位的影响机制，并提出了基于阵列弯曲结构特性的相位解模糊处理思路。建立了频率差分的等效阵列模型，并且针对阵列参数误差建立了分布式阵列稳健方位估计方法。主要完成了（1）基于频率差分方法的空间抗混叠方位估计方法研究，可实现阵元间距是标准半波长阵元间距4倍的分布式阵列抗混叠方位估计，并得到海上实测数据验证；（2）严重弯曲变形阵列自校准方位估计技术研究，针对SWellEx96公开数据集，孔径100米的海底固定水平阵阵元位置估计误差不大于2米；（3）阵元位置与响应误差并存时的方位估计方法研究，可实现传感器幅度、相位、阵元位置以及目标方位参数的联合估计；（4）海洋非均匀环境噪声下的稳健方位估计方法研究，可实现浅海柱面非各向同性噪声场条件下目标方位的准确估计。研究成果共发表SCI论文6篇，申请专利1项，授权专利3项。