基于IVMD-RPE的船舶噪声信号去噪及特征提取方法  

Ship signal denoising and feature extraction based on IVMD-RPE

作　　者：刘夏扬 李晶 赵国新[1] 刘昱 雷琦 LIU Xiayang;LI Jing;ZHAO Guoxin;LIU Yu;LEI Qi(College of Information Engineering,Beijing Institute of Petrochemical Technology,Beijing 102617,China)

机构地区：[1]北京石油化工学院信息工程学院,北京102617

出　　处：《声学技术》2023年第3期297-305,共9页Technical Acoustics

基　　金：国家自然科学基金青年基金(51405023)。

摘　　要：针对由于复杂环境条件下的船舶噪声信号识别度低的问题,提出了一种改进的变分模式分解(Improved Variational Mode Decomposition,IVMD)、反向排列熵(Reverse Permutation Entropy,RPE)、加权排列熵(Weighted Permutation Entropy,WPE)和能量比相结合的船舶噪声信号识别分类方法。该方法利用IVMD将信号分解为若干本征模态函数(Intrinsic Mode Function,IMF);再利用RPE对IMF进行筛选,得到敏感IMF,实现去噪过程;最后计算IMF的WPE并与各个IMF的能量比构建特征向量,建立长短期记忆网络(Long Short-Term Memory,LSTM)进行识别分类。实验结果表明,与经验模态分解相比,文中提出的基于IVMD-RPE的船舶噪声信号去噪及特征提取方法能有效减少环境噪声的影响,提高信噪比,对船舶噪声目标信号识别分类的准确率更高。Aiming at the practical problem of low recognition degree of ship-radiated noise under complex environmental conditions,a method combining improved variational mode decomposition(IVMD),reverse permutation entropy(RPE),weighted permutation entropy(WPE)and energy ratio is proposed for ship noise signal recognition and classification.In this method,IVMD is used to decompose the signal into several intrinsic mode functions(IMFs),RPE is used to screen IMFs to obtain sensitive IMFs and realize the denoising process,and WPE of IMF is calculated and compared with the energy ratio of each IMF.Finally,the feature vector is constructed and the long short-term memory(LSTM)network is established for recognition and classification.The experimental results show that,compared with the empirical mode decomposition,the proposed method based on IVMD and RPE can effectively reduce the influence of environmental noise,improve the signal-to-noise ratio,and have a higher accuracy in the identification and classification of ship noise signals.

关 键 词：变分模式分解(VMD) 反向排列熵(RPE) 长短期记忆网络(LSTM) 

分 类 号：TM911.7[电气工程—电力电子与电力传动]