深度学习技术的发展为船舶辐射噪声分类识别提供了一个新的方法。本文从人耳听觉角度出发，提出一种融合人耳听觉特性与堆栈自编码神经网络的船舶辐射噪声分类方法。该方法使用Mel滤波器模拟人耳对噪声信号频率的选择，借助SAE网络逐层自动提取舰船辐射噪声人耳听觉特征量的深度特征，并将该特征用于分类识别。针对实测船舶辐射噪声信号进行试验，结果表明，本文提出的方法具有91.19%的识别正确率。
The development of deep learning technology provides a new method for recognition of ship radiated noise. From the point of view of human ear auditory characteristics, this paper proposed a classification method of ship radiation noise which combines human ear auditory characteristics and stack self-coding neural network. In the method, Mel filter is used to simulate the selection of noise signal’s frequency by human ear, and SAE model is used to automatically extract the depth characteristics of human ear auditory characteristics of ship radiation noise. After that, the depth characteristics are used for recognition. Through the tests of the measured ship radiated noise, it is confirmed that the method proposed in this paper has 91.19% recognition accuracy.
2020,42(8): 172-176 收稿日期：2020-03-12
DOI：10.3404/j.issn.1672-7649.2020.08.032
分类号：TP183
作者简介：李海涛(1988-),男,博士,研究方向为水声目标识别
参考文献：
[1] SABARA R, JESUS S. Underwater acoustic target recognition using graph convolutional neural networks[J]. The Journal of the Acoustic Society of America, 2018, 144(3): 1744-1757
[2] KAMAL S, MOHAMMED S K, PILLAI P R S, et al. Deep learning architectures for underwater target recognition[J]. Ocean Electronics, 2013: 48-54.
[3] YU Y, CAO X, ZHANG X M. Underwater target classification using deep neural network[C]. Oceans. IEEE, 2018.
[4] WEI Z X, JU Y, SONG M. A method of underwater acoustic signal classification based on deep neural network[C]. ICISCE, 2018.
[5] 刘振, 邱家兴, 程玉胜. 一种基于深度神经网络的船舶辐射噪声分类方法[C]. 2018年鲁浙苏黑四省声学技术学术交流会论文集, 2018.
[6] WU H, SONG Q Z, JIN G H. Deep learning based framework for underwater acoustic signal recognition and classification[C]. CSAI, 2018: 385-388.
[7] WANG N B, HE M, SUN J G. ia-PNCC: Noise processing method for underwater target recognition convolutional neural network[J]. CMC-Computers Materials & Continua, 2019, 58(1): 169-181
[8] KE X Q, YUAN F, CHENG E. Underwater acoustic target recognition based on supervised feature-separation algorithm[J]. Sensors, 2018, 18(12): 4318-4341
[9] CAO X, TOGNERI R, ZHANG X, et al. Convolutional neural network with second-order pooling for underwater aarget classification[J]. IEEE Sensors Journal, 2018, 19(8): 3058-3066
[10] SHEN S, YANG H, SHENG M. Compression of a deep competitive network based on mutual information for underwater acoustic targets recognition[J]. Entropy, 2018, 20(4): 243-255
[11] HU Gang, WANG Kejun, PENG Yuan, et al. Deep learning methods for underwater target feature extraction and recognition[J]. Computational Intelligence & Neuroscience, 2018(3): 1-10
[12] 赵增科. 基于深度学习的水下目标识别[D]. 哈尔滨: 哈尔滨工程大学, 2017.
[13] 王强, 曾向阳. 深度学习方法及其在水下目标识别中的应用[C]. 中国声学学会水声学分会2015年学术会议论文集. 2015.
[14] CHEN Y, XU X. The research of underwater target recognition method based on deep learning[C]. IEEE International Conference on Signal Processing, Communications and Computing, 2017: 1-5.
[15] 杨宏晖, 申昇, 姚晓辉, 等. 用于水声目标特征学习与识别的混合正则化深度置信网络[J]. 西北工业大学学报, 2017, 35(2): 220-225
[16] YANG H, SHEN S, YAO X, et al. Competitive deep-belief Networks for underwater acoustic target recognition.[J]. Sensors, 2018, 18(4): 951-964
[17] 陈越超, 徐晓男. 基于降噪自编码器的水中目标识别方法[J]. 声学与电子工程, 2018(1): 30-33
[18] CAO Xu, ZHANG Xiaomin, YU Yang, et al. Deep learning-based recognition of underwater target[C]. IEEE International Conference on Digital Signal Processing, 2017: 89-93.
[19] 曹旭, 张效民, 于洋, 等. 基于深度学习的水下目标识别方法[P]. 中国: CN106529428A, 2017.
[20] 张大伟, 章新华, 付留芳, 等. 基于听觉感知与卷积神经网络识别舰船目标[C]. 中国声学学会青年学术会议会议, 2015.
[21] YANG H H, LI J H, SHEN S, et al. A deep convolutional neural network inspired by auditory perception for underwater acoustic target recognition[J]. Sensors, 2019, 19(5): 1104-1115
[22] GOODFELLOW I, BENGIO Y, COURVILLE A. Deep learning[M]. Mit Press, 2016: 94-95.
[23] 王嵘冰, 徐红艳, 李波, 等. BP神经网络隐含层节点数确定方法研究[J]计算机技术与发展, 2018, 28(4): 31-35.
