一种基于典型相关分析的主动声纳水下混响抑制算法

混响抑制是提升主动声纳接收机性能的关键技术之一。传统的混响抑制算法主要采用多普勒和时间增益控制,以及预白化处理。这些算法需要对混响建模并进行参数估计,其计算量较大,且易受水声环境变化的影响。为了提升抗混响算法的鲁棒性,提出了一种基于典型相关分析的算法,能够与支持向量机结合实现对混响的预先分类识别。湖上试验结果表明该算法能够显著提升混响背景下主动声纳的检测性能。

CCA- based Approach to Improve Active Sonar Detection in Reverberation

Reverberation suppression is one of the most important approaches to improve active sonar detection performance. One can suppress reverberation using several traditional algorithms,such as Doppler control and TGC,or prewhitening methods. Traditional methods need to set up reverberation model and make parameter estimation,which require large computation cost,and the results are deeply influenced by underwater environment. A canonical correlation analysis（ CCA） algorithm is used to make reverberation suppres- sion much more robust,which can classify reverberation with the help of support vector machine（ SVM）. From the numerical analysis of lake experiment data,the proposed method is verified to be efficient to improve active sonar detection performance in reverberation.