Open Access
 Issue JNWPU Volume 37, Number 1, February 2019 87 - 92 https://doi.org/10.1051/jnwpu/20193710087 03 April 2019

© 2019 Journal of Northwestern Polytechnical University

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

## 2 MKSPMCCA特征融合算法

KSPCCA算法原理上只能是对2组特征进行处理, 但是水下目标识别要对多组特征进行融合。因此本文提出了MKSPMCCA算法。该算法通过对多域特征进行融合获得更有效的判别信息。

α(i)=ϕif(X(i))β(i), 可得到如下优化模型

1) 提取每类水下目标的多域特征构成特征样本集X(i)(i=1, 2, …, m);

2) 利用核稀疏保持投影算法对样本空间进行核稀疏化, 求多特征集X(i)的多核稀疏表示系数sn(i)(i=1, 2, …, m; n=1, 2, …, N), 构造多核稀疏重构邻接矩阵S(i)=[s1(i), …, sN(i)];

3) 求解模型(4)式将得到的特征值按照由大到小的顺序进行排序, 选择前d个最大的特征值, 并计算出相应的特征向量βj(i)(i=1, 2, …, m; j=1, 2, …, d);

4) 计算每组样本特征集的投影向量α(i);

## 3 基于不同特征融合算法的水下目标识别性能实验研究

1) 实验数据

2) 实验流程

### 3.2 引入核方法前后典型相关变量对确定实验

 图1典型相关系数
 图2方差贡献率和累积方差贡献率

### 3.3 基于2组特征融合的CCA和KSPCCA算法的对比实验

 图3特征两两融合的识别结果

### 3.4 基于多组特征融合的MCCA和MKSPMCCA算法的对比实验

 图4多特征融合的识别结果

## References

1. Meng Q, Yang S, Piao S. The classification of Underwater Acoustic Target Signals Based on Wave Structure and Support Vector Machine[J]. Journal of the Acoustical Society of America, 2014, 136(4): 2265-2265 [Article] [CrossRef] [Google Scholar]
2. Cheng Yusheng, Zhang Baohua, Gao Xing, et al. Phase-Coupling Characteristics of Ship Radiiated-Noise Demodulation Spectrum And Application[J]. Acta Acustica, 2012(1): 25-29 (in Chinese) [Article] [Google Scholar]
3. Jiao Y M, Kang C Y, Zeng X X, et al. Extraction and Application in Nonlinear Spectrum Feature of Ship Radiated Noise[J]. Ship Science & Technology, 2016, 38(12): 65-68 [Article] [Google Scholar]
4. Slamnoiu G, Radu O, Rosca V, et al. DEMON-Type Algorithms for Determination of Hydro-Acoustic Signatures of Surface Ships and of Divers[C]//Materials Science and Engineering Conference Series, 2016 [Google Scholar]
5. Liu Y, Zhang X, Shao J. Quadratic Time-Frequency Feature Extraction and Fusion for Ship Targets Classification[C]//International Conference on Signal Processing, 2015 [Google Scholar]
6. Cao H L, Fang S L, Luo X W. Nonlinear Feature Extraction and Recognition of Ship Radiated Noise[J]. Journal of Nanjing University, 2013, 49(1): 64-71 [Article] [Google Scholar]
7. Yang H, Shen S, Yao X, et al. Competitive Deep-Belief Networks for Underwater Acoustic Target Recognition[J]. Sensors, 2018, 18(4): 952 [Article] [CrossRef] [Google Scholar]
8. 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 [Article] [CrossRef] [Google Scholar]
9. Yuan Shuai, Yang Honghui, Shen Xing. Forward Order Feature Selection Algorithm Based on Mutual Information[J]. Technical Acoustics, 2014(4): 359-362 (in Chinese) [Article] [Google Scholar]
10. Ma Chao, Chen Xihong, Xu Yuliang, et al. Ensemble Feature Selection Based on Generalized Neighborhood Rough Model and Its Selective Integration[J]. Journal of Xi'an Jiaotong University, 2011, 45(6): 34-39 (in Chinese) [Article] [Google Scholar]
11. Yang Honghui, Dai Jian, Sun Jincai, et al. A New Adaptive Immune Feature Selection Algorithm for Underwater Acoustic Target Classification[J]. Journal of Xi'an Jiaotong University, 2011, 45(12): 28-32 (in Chinese) [Article] [Google Scholar]
12. Liu Junfeng, Zhang Xinhua, Xu Linzhou. Application of Dynamic Programming in Passive Sonar for Detecting Target[J]. Ship Science & Technology, 2012, 34(3): 95-98 (in Chinese) [Article] [Google Scholar]

## All Figures

 图1典型相关系数 In the text
 图2方差贡献率和累积方差贡献率 In the text
 图3特征两两融合的识别结果 In the text
 图4多特征融合的识别结果 In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.