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  

Feature line spectrum extraction of ship radiated noise based on variable mode decomposition optimized by genetic algorithm

作　　者：沈鑫玉 陈涛[3] 郭良浩[1] 刘建军[1] 陈艳丽 SHEN Xinyu;CHEN Tao;GUO Lianghao;LIU Jianjun;CHEN Yanli(State Key Laboratory of Acoustics,Institute of Acoustics,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China;91198 Troops of the Chinese People’s Liberation Army,Guangxi 536122,China)

机构地区：[1]中国科学院声学研究所声场声信息国家重点实验室,北京100190 [2]中国科学院大学,北京100049 [3]中国人民解放军91198部队,广西536122

出　　处：《应用声学》2024年第1期1-11,共11页Journal of Applied Acoustics

摘　　要：特征线谱提取是舰船目标识别的一个重要研究环节,常采用传统的DEMON谱分析方法,处理过程中,一般对舰船噪声时域信号未予抑噪,低信噪比情况下,传统DEMON谱分析性能差。对此,提出一种采用遗传算法优化变分模态分解方法,用于分解舰船噪声原时域信号,获得抑制噪声后的舰船噪声重构信号,进而有效提取了舰船目标噪声幅度调制特征线谱。该方法首先采用遗传算法优化变分模态分解的两个关键输入参数(分解所取模态个数和惩罚因子),对变分模态分解得到的各阶固有模态分量加以判别,去除噪声主导分量,保留信号主导分量,使重构舰船噪声信号显著抑制了干扰噪声,然后对降噪后的重构信号进行频谱分析,获得目标噪声调制特征线谱。理论分析、仿真和实验数据处理结果表明,相比传统DEMON谱分析法,基于遗传算法优化变分模态分解的舰船噪声特征线谱提取方法具有更好的噪声抑制能力,所获取的舰船噪声幅度调制特征线谱信噪比明显高于传统DEMON方法,具有一定优势,前景良好。Feature line spectrum extraction is an important part of ship target recognition.The traditional detection of envelope modulation on noise(DEMON)spectral analysis method is often used.In the process of processing,the time-domain signal of ship noise is generally not suppressed.In the case of low signal to noise ratio(SNR),the performance of traditional DEMON spectral analysis is poor.In this paper,a genetic algorithm(GA)optimized variational mode decomposition(VMD)method is proposed to decompose the original time-domain signal of ship noise,obtain the reconstructed signal of ship noise after noise suppression,and then effectively extract the amplitude modulation characteristic line spectrum of ship target noise.Firstly,GA is used to optimize the two key inputs of VMD(the number of modes and penalty factor),decompose the VMD to obtain the intrinsic mode function(IMF)of each order,judge,remove the dominant component of noise,and retain the dominant component of signal,so that the reconstructed ship noise signal can significantly suppress the interference noise.Then the target noise modulation characteristic line spectrum is obtained by spectral analysis of the reconstructed signal after noise reduction.The theoretical analysis,simulation and experimental data processing results show that compared with the traditional DEMON spectral analysis method,the GA-VMD based ship noise feature line spectrum extraction method has better noise suppression ability,and the obtained ship noise amplitude modulation feature line spectrum SNR is significantly higher than the traditional DEMON method,which has certain advantages and good prospects.

关 键 词：舰船辐射噪声 遗传算法 变分模态分解 特征线谱提取 

分 类 号：TN911.7[电子电信—通信与信息系统]