基于鲁棒主成分分析的低频水声信号降噪方法  

Noise reduction method of low frequency underwater acoustic signal based on robust principal component analysis

作　　者：吴耀文 邢传玺[1] 岳露露 万兴举 WU Yao-wen;XING Chuan-xi;YUE Lu-lu;WAN Xin-ju(School of Electrical and Information Technology,Yunnan Minzu University,Kunming 650500,China)

机构地区：[1]云南民族大学电气信息工程学院,云南昆明650500

出　　处：《云南民族大学学报（自然科学版）》2020年第1期70-77,共8页Journal of Yunnan Minzu University:Natural Sciences Edition

基　　金：国家自然科学基金(61761048);云南省高校科技创新团队支持计划.

摘　　要：由于海洋环境噪声的复杂性,接收到的信号往往具有较低的信噪比,导致水声信号处理难度大等问题.针对此问题,采用基于鲁棒主成分分析的降噪方法,建立将含噪信号表示为低秩、稀疏和噪声的分解模型,研究了对低频水声信号的降噪问题.首先通过Godec算法将含噪信号表示为低秩、稀疏和噪声3部分,然后运用非负矩阵分解算法对低秩部分进行分解,得到噪声字典,最后根据得到的噪声字典从含噪信号中提取出初始水声信号.通过对不同海况下即不同信噪比的仿真信号进行降噪处理,结果表明该方法在水声信噪分离中具有较好的降噪效果.Due to the complexity of marine environment noise,the received signal often has a low signal-to-noise ratio,which leads to the difficulty of low frequency underwater acoustic signal processing and so on.In order to solve this problem,a low rank,sparse and noisy decomposition model is established based on robust principal component analysis.Firstly,the noisy signal is decomposed into three parts by Godec algorithm:low rank,sparse and noise,and then the low rank part is learned by non-negative matrix decomposition,and the noise dictionary is obtained.Finally,the pure underwater acoustic signal is extracted from the noisy signal by using the obtained noise dictionary.The results show that the method has better noise reduction effect in the underwater acoustic signal-noise separation by noise reduction processing of the simulation signals with different signal-to-noise ratios under different sea conditions.

关 键 词：稀疏低秩分解 非负矩阵分解 字典学习 WIENER滤波 

分 类 号：TN929.3[电子电信—通信与信息系统]