目标线谱提取是舰船识别的重要研究对象，常用的DEMON谱分析法，通常由于有色噪声的干扰而失效。针对该问题，本文基于线谱在频域上的稀疏特性，提出一种关于辐射噪声信号的改进Lasso回归算法，用以抑制有色背景噪声，凸显出目标线谱特征。首先，通过分析噪声将信号转化为可分析的数学模型，从而构造出Lasso中的残差项；再根据线谱剔除趋势相对于零点的不同稀疏度，以及线谱自身的稀疏特性构造出Lasso中的惩罚项。理论推导、仿真及海试数据表明，相比常用的DEMON谱分析，该方法在–15 dB的有色噪声干扰情况下能抑制大部分噪声，最大限度地保留目标线谱特征，具有一定的应用价值。
Target line spectrum extraction is an important research object of ship identification, and the commonly used DEMON spectrum analysis method usually fails due to the interference of colored noise. In order to solve this problem, based on the sparse characteristics of the line spectrum in the frequency domain, an improved Lasso regression algorithm for radiated noise signals is proposed to suppress the colored background noise and highlight the characteristics of the target line spectrum. Firstly, the signal was transformed into an analyzable mathematical model by analyzing the noise, so as to construct the residual term in Lasso. Then, according to the different sparsity of the line spectrum elimination trend relative to the zero point, and the sparsity characteristics of the line spectrum itself, the penalty term in Lasso is constructed. Theoretical derivation, simulation and sea trial data show that compared with the commonly used DEMON spectrum analysis, the proposed method can suppress most of the noise and retain the target line spectrum characteristics to the greatest extent under the interference of -15 dB colored noise, which has certain application value.
2025,47(10): 130-137 收稿日期：2024-6-11
DOI：10.3404/j.issn.1672-7649.2025.10.022
分类号：TN911.7
基金项目：国家重点研发计划项目（2023YFB3907202）；青岛哈尔滨工程大学创新发展中心青年科学家培育基金项目（79000013/007）
作者简介：罗淦（1997-），男，硕士研究生，研究方向为自动控制及水下通信
参考文献：
[1] 刘伯胜, 雷家煜. 水声学原理(第二版)[M]. 哈尔滨: 哈尔滨工程大学出版社, 2009.
[2] 王洪玲. 舰船辐射噪声调制特征提取方法研究[D]. 哈尔滨: 哈尔滨工程大学, 2007.
[3] JANSEN E, JONG C D. Experimental assessment of underwater acoustic source levels of different ship types[J]. IEEE Journal of Oceanic Engineering, 2017, 42(2): 1-10.
[4] ROSENLICHT M. Introduction to spectral analysis[M]. New York: Dover Publications, 2005.
[5] FLORESCU A, CIOCHINA S. Refining accuracy of the spectral lines estimation by a sparsity based approach[C]//Proceedings of 9th International Conference on Communications, Bucharest, 2012.
[6] CHEN Z F, LI J, TAN X, et al. On probing waveforms and adaptive receivers for active sonar[J]. Oceans 2010 MT S/IEEE SEATTLE, Seattle USA, 2010: 1–10.
[7] TAN X, ROBERTS W, LI J, et al. Sparse learning via iterative minimization with application to MIMO radar imaging[J]. IEEE Transactions on Signal Processing, 2011, 59(3): 1088-1101.
[8] 张苗辉, 李晖, 陈宁, 等. 一种改进的中值滤波方法进行背景估计[J]. 江西科学, 2018, 36(2): 308-313.
ZHANG M H, LI H, CHEN N, et al. An improved median filtering method for background estimation[J]. Jiangxi Science, 2018, 36(2): 308-313.
[9] 周武, 张宏滔. 水下无人航行器自主检测方法研究[J]. 声学技术, 2020, 39(2): 146-150.
ZHOU W, ZHANG H T. Research on autonomous detection methods for underwater unmanned vehicles[J]. Ac oustic Technology, 2020, 39(2): 146-150.
[10] 张哲. 高维数据线性回归建模方法分析[D]. 天津: 天津大学, 2013.
[11] 陶笃纯. 螺旋桨空化噪声谱[J]. 声学学报, 1982, 7(6): 344-351.
TAO D C. Propeller cavitation noise spectrum[J]. Acta Acoustica Sinica, 1982, 7(6): 344-351.
[12] 蒋国健. 舰船螺旋桨空泡噪声的数理模型[J]. 声学学报, 1998, 23(5): 401–408.
JIANG G J. Mathematical model of ship propeller cavitation noise[J]. Acta Acoustica Sinica, 1998, 23(5): 401–408.
[13] 孙军平, 林建恒,杨军, 等. 准周期随机声脉冲序列信号在浅海波导中的传播[J]. 声学学报, 2017, 42(1): 21–26.
SUN J P, LIN J H, YANG J, et al. Propagation of periodic randomacoustic pulse sequence signal in shallow sea waveguide[J]. Acta Acoustica Sinica, 2017, 42(1): 21–26.
[14] 陶笃纯. 舰船噪声节奏的研究(Ⅰ)–数学模型及功率谱密度[J]. 声学学报, 1983(2): 65-76.
TAO D C. Research on ship noise rhythm (I)–Mathematical model and power spectral density[J]. Acta, 1983(2): 65-76.
[15] 沈鑫玉, 陈涛, 郭良浩, 等. 遗传算法优化变分模态分解提取舰船辐射噪声特征线谱方法[J/OL]. 应用声学: 1–14 [2024-03-15].
SHEN X Y, CHEN T, GUO L H, et al. Genetic algorithm optimized variational mode decomposition method f or extracting characteris tic line spectra of ship radiated noise [J/OL]. Applied Acoustics: 1–14 [2024-03-15].
[16] 刘子昂, 蒋雪, 伍冬睿. 基于池的无监督线性回归主动学习[J]. 自动化学报, 2021, 47(12): 2771-2783.
LIU Z A, JIANG X, WU D R. Unsupervised linear regression active learning based on pools[J]. Journal of Automation, 2021, 47(12): 2771-2783.
[17] SELESNICK I W, ARNOLD S, DANTHAM V R. Polynomial smoothing of time series with additive step discontinuities[J]. IEEE Transactions on Signal Process, 2012, 60(12): 6305–6318.
[18] NING X R, IVAN W S, LAURENT D. Chromatogram baseline estimation and denoising using sparsity (BEADS)[J]. Chemometrics and Intelligent Laboratory Systems, 2014, 139: 156–167.
[19] CANDES E J, TAO T. Decoding by linear programming[J]. IEEE Transactions on Information Theory, 2006, 51(12): 4203-4215.
