基于目标辐射噪声特征的检测新技术研究

Conventional passive sonar detection is generally based on wide-band energy integral. The spectral characteristic of target radiated noise is not fully considered. Due to the propeller rotation or other mechanical vibrations, target radiated noise contains a lot of line spectrum ingredients and strong narrowband signal, even strong broadband ingredients in certain frequency band. In this project， according to the common target radiated noise characteristics and ambient noise characteristics，signal processing band is reasonably split into different frequency bands(frequencies). Then, in order to gain the directivity and the gain of different frequency band signals in the broadband, the project researches the best lineup of unequal spacing in the condition of certain number of array elements. Finally, the fusion method of the processing results of different frequency bands (frequency) is researched to achieve effective detection of target. The study fully taking advantage of the relationship between the input SNR and processing gain which are closely related with the band , can improve the output SNR, so as to improve the detection performance. The research results based on target radiated noise detection method will provide a new mechanism of passive sonar technology, and leap to a new level of passive sonar performance in China.

常规被动声纳检测一般是基于宽带能量积分的，没有充分考虑目标辐射噪声的频谱特性。由于螺旋桨旋转或其他机械振动，在目标辐射噪声中含有丰富的线谱成份和较强的窄带成份，甚至在某些频带上出现较强的宽带成份。本项目拟根据常见目标辐射噪声特性与环境噪声特性的差异，对信号处理频带进行最佳频带（频率）分割，然后在此基础上从满足宽带内不同频带信号指向性和增益要求的角度，研究一定阵元数条件下，不等间距最佳布阵方式，最后研究对不同频带（频率）处理结果的融合方法，实现对目标的有效检测。该项研究很好地利用与频带密切相关的输入信噪比和处理增益的关系，提高信号处理输出信噪比，从而提高检测性能。基于目标辐射噪声的检测新方法的研究成果将为被动声纳技术提供新机理，使我国被动声纳性能跃上一个新台阶。

常规被动声呐检测一般是基于宽带能量积分的，没有充分考虑目标辐射噪声的频谱特性。由于螺旋桨旋转或其他机械振动，在目标辐射噪声中含有丰富的线谱成份和较强的窄带成份。项目根据目标辐射噪声特性与海洋环境噪声特性的差异，对信号处理频带进行最佳频带（频率）分割，在此基础上从满足宽带内不同频带信号指向性和增益要求的角度，研究一定阵元数条件下，不等间距最佳布阵方式。项目利用不等间距阵进行湖上、海上试验，对不同频带（频率）信号进行精细分析，对分析结果进行数据融合。项目着重对目标辐射噪声中线谱频率未知的弱目标探测算法进行研究，根据辐射噪声的空、时稳定性特征及各频带信号分布特性提出信号处理算法，提高信号处理输出信噪比，从而实现弱目标探测。本项目历时4年，与国内、外学者开展了广泛且深入的学术交流，培养硕士、博士研究生5名，发表文章、专利等多份，其中8篇文章被EI检索，4篇文章被SCIE/SSCI检索。