稳健的矢量阵被动测向算法研究

The acoustic vector sensor is combined by omnidirectional pressure sensor and dipole particle velocity sensor. The acoustic vector sensor can obtain the unambiguous direction of the signal source. Under the condition of the same number of sensors, the vector-sensor array outperforms the scalar-sensor array. .However, the above-mentioned advantages of the vector-sensor array is obtained based on the assumption that the received signal is stationary. In practice, the radiation signal from the submarine passes through the complex ocean environment, and the radiation signal impinging on the vector sensor is non-stationary. Thus, the vector sensor can obtain only a small number of snapshots during the signal stationary period. In addition, in the practical war environment, the objectives are fast-moving targets or instantons signal, therefore, the vector sensor only obtains a small number of meaningful snapshots..It is noticed that most of existing statistic-based algorithms are no longer applicable to a small number of snapshots. On the other hand, the vector sensor obtains the theoretical performance in absence of channel gain-phase errors. In the practical environment, the channel gain-phase errors are unavoidable. Most of high spatial resolution algorithms are sensitive to channel gain-phase errors, especially phase errors..In order to improve the feasibility of the vector-sensor array in the practical application, this project focus on three tasks :1) develop the vector-sensor array method for a small number of snapshots; 2) develop the robust vector-sensor array method against the channel gain-phase errors; 3) develop the channel gain-phase error self-calibration method.

矢量水听器由传统的声压水听器和质点振速水听器复合而成，能在全空间对声源进行无模糊定向。多个矢量水听器组成的声纳基阵能获得比相同数目的声压阵更好的性能。上述优良性能都是假设矢量水听器的入射信号是平稳的。实际中潜艇的辐射信号经过复杂的海洋环境信道到达矢量水听器时往往是一个非平稳过程，此时在信号平稳周期内矢量水听器获得的快拍数目有限。另外，实际作战环境中, 大量出现的是高速运动目标, 此时获取的有意义数据快拍较少。基于统计量的传统算法不能对小快拍数据有效测向。另外，矢量水听器的理论性能都是在通道幅相误差为零的假设下得到的。在实际应用中，通道幅相误差不可避免。大部分的高分辨算法对于通道幅相误差特别是相位误差较为敏感。.因此，为了提高矢量阵在实际复杂海洋环境中的适用性，本项目研究如下内容: 1) 小快拍数据下矢量阵被动测向算法; 2) 对通道幅相误差稳健的测向算法; 3) 矢量阵幅相误差自校正算法。

本项目预定的研究内容为矢量阵在复杂海洋环境下(幅相误差和小快拍等)的稳健测向与自校正方法。本项目顺利完成此研究内容，提出了对幅相误差和小快拍稳健的阵列测向和自校正方法。另外，针对单矢量传感器，提出了单目标情况下，对相位误差完全免疫的稳健测向和自校正方法。在研究过程中，我们注意到矢量传感器声压和振速通道之间存在噪声功率不一致性，严重影响其在低信噪比情况下的测向性能。根据矢量传感器的特性，提出简单有效的虚源消除方法。考虑到噪声功率不一致情况下的信源数估计难题，提出一种基于遗传算法和MDL准则的信源数和方位估计方法。这些研究工作进一步提高了矢量阵在水下环境中的测向稳健性。截至目前，参加相关国际学术会议2次，培养博士生和硕士生各2人；在国内外知名期刊上共发表和录用项目资助论文9篇，其中，SCI检索论文5篇，EI检索论文4篇。SCI论文web of science 总计引用14次，EI中文期刊论文CNKI引用总计4次。以本项目资助申请相关专利5项。