一种基于脑网络特征的水声目标识别算法

doi:10.12382/bgxb.2024.0735

摘要/Abstract

摘要：

针对声呐员在水声目标识别过程中脑力负荷大、无法保证长时间有效工作状态的问题,基于脑-机接口技术,提出一种基于脑网络特征的水声目标识别算法,用于辅助声呐员完成水下目标的快速识别。为了增强模型对大脑神经活动信息的提取,并降低大脑无关依赖性的干扰,利用格兰杰因果和转移熵理论重建脑网络特征提取算法,并将其用于水声目标分类模型的构建。设计视-听联合刺激范式模拟真实工作环境并进行实验数据采集,以完成水声目标分类模型的训练与验证。分析结果表明,新提出的脑网络特征算法可以更好地捕获神经活动中的依赖性信息,结合所设计的视-听联合刺激范式,完成了对基于脑网络特征的水声目标分类模型验证实验,最终识别准确率稳定在90%以上。

关键词: 脑-机接口, 水声目标识别, 脑电图, 脑网络, 支持向量机

Abstract:

In response to the problems of sonar operator having a heavy mental workload and the inability to ensure long-term effective working status in the process of underwater target recognition, a brain network feature-based underwater target recognition algorithm based on brain-computer interface (BCI) technology is proposed to assist sonar operators in achieving the rapid recognition of underwater targets. In order to enhance the extraction of brain neural activity information by the model and reduce the interference of brain irrelevant dependencies, the Granger causality (GC) and transfer entropy (TE) theories are used to reconstruct a brain network feature extraction algorithm, and a underwater acoustic target classification model is established by the proposed algorithm. A visual-auditory joint stimulation paradigm is designed for environmental simulation, and the experimental data is collected to complete the training and validation of the underwater acoustic target classification model. The analyzed results show that the proposed brain network feature algorithm can better capture the dependency information in neural activity. The validation of the underwater acoustic target classification model based on brain network features is verified by the visual-auditory joint stimulation paradigm, and the final recognition accuracy is over 90%.

Key words: brain-computer interface, underwater acoustic target recognition, electroencephalogram, brain network, support vector machine

中图分类号:

R338

张家琦, 石章松, 徐慧慧. 一种基于脑网络特征的水声目标识别算法[J]. 兵工学报, 2025, 46(5): 240735-.

ZHANG Jiaqi, SHI Zhangsong, XU Huihui. An Underwater Acoustic Target Recognition Algorithm Based on Brain Network Features[J]. Acta Armamentarii, 2025, 46(5): 240735-.

图/表 12

图1 水声目标识别整体流程图

Fig.1 Overall flowchart of underwater target recognition

图2 信号处理流程图

Fig.2 Signal processing flowchart

图3 实验范式流程图

Fig.3 Experimental paradigm flowchart

图4 EEG信号采集及实验装置图

Fig.4 EEG signal acquisition and experimental setup

图5 EEG信号预处理前后对比图

Fig.5 Comparison of EEG signals before and after preprocessing

图6 视听刺激信号与EEG信号对比图

Fig.6 Comparison of audio-visual stimulation signals and EEG signals

表1 不同时刻任务下脑网络特征一致性检验结果

Table 1 Consistency test results of brain network features under different time tasks

试次对	龙虾		海豚		蝠鲼
试次对	加权 kappa值	显著性	加权 kappa值	显著性	加权 kappa值	显著性
1-2	0.828	<0.001	0.832	<0.001	0.825	<0.001
1-3	0.809	<0.001	0.825	<0.001	0.829	<0.001
1-4	0.839	<0.001	0.833	<0.001	0.828	<0.001
1-5	0.795	<0.001	0.837	<0.001	0.866	<0.001
1-6	0.829	<0.001	0.813	<0.001	0.821	<0.001
2-3	0.841	<0.001	0.829	<0.001	0.830	<0.001
2-4	0.854	<0.001	0.870	<0.001	0.855	<0.001
2-5	0.843	<0.001	0.855	<0.001	0.845	<0.001
2-6	0.845	<0.001	0.838	<0.001	0.811	<0.001
3-4	0.840	<0.001	0.833	<0.001	0.828	<0.001
3-5	0.819	<0.001	0.852	<0.001	0.877	<0.001
3-6	0.831	<0.001	0.825	<0.001	0.849	<0.001
4-5	0.864	<0.001	0.857	<0.001	0.853	<0.001
4-6	0.813	<0.001	0.828	<0.001	0.831	<0.001
5-6	0.806	<0.001	0.807	<0.001	0.851	<0.001

图7 1号受试者组平均GT脑网络对比

Fig.7 Comparison of group average GT brain networks of Subject 1

图8 45名受试者在不同脑网络模型下分类准确率分布图

Fig.8 Distribution of classification accuracies among 45 subjects under different brain network models

图9 不同特征集维度下45名受试者的平均分类准确率

Fig.9 Average classification accuracy of 45 subjects under different feature set dimensions

表2 三类脑网络模型分类结果

Table 2 Classification results of three types of brain network models

算法	特征集维度	测试准确率/%	标准差
GT	185	87.58	2.73
GC	302	68.84	8.59
TE	203	56.23	9.26

表3 不同特征提取模型下的分类准确率

Table 3 Classification accuracies under different feature extraction models

算法	测试准确率/%	标准差
脑网络+LSTM	94.16	0.0284
MFCC+LSTM	84.93	0.0308

参考文献 35

[1]	MÜLLER N, REERMANN J, MEISEN T. Navigating the depths: a comprehensive survey of deep learning for passive underwater acoustic target recognition[J]. IEEE Access, 2024, 12: 154092-154118.
[2]	DOAN V S, HUYNH-THE T, KIM D S. Underwater acoustic target classification based on dense convolutional neural network[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19: 1-5.
[3]	LEI F, TANG F F, LI S H. Underwater target detection algorithm based on improved YOLOv5[J]. Journal of Marine Science and Engineering, 2022, 10(3):310.
[4]	XU G F, ZHOU D X, YUAN L B, et al. Vision-based underwater target real-time detection for autonomous underwater vehicle subsea exploration[J]. Frontiers in Marine Science, 2023, 10: 1112310.
[5]	薛灵芝, 曾向阳, 杨爽. 基于生成对抗网络的水声目标识别算法[J]. 兵工学报, 2021, 42(11):2444-2452. doi: 10.3969/j.issn.1000-1093.2021.11.018
	XUE L Z, ZENG X Y, YANG S. Underwater acoustic target recognition algorithm based on generative adversarial networks[J]. Acta Armamentarii, 2021, 42(11): 2444-2452. (in Chinese) doi: 10.3969/j.issn.1000-1093.2021.11.018
[6]	MENG Q X, YANG S E, PIAO S C. 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.
[7]	MOURA N N D, SEIXAS J M D. Novelty detection in passive SONAR systems using support vector machines[C]//Proceedings of 2015 Latin America Congress on Computational Intelligence.Washington,D.C., US:IEEE, 2015.
[8]	DAVID S D, SOLEDAD T G, ANTONIO C L, et al. ShipsEar: an underwater vessel noise database[J]. Applied Acoustics, 2016, 113: 64-69.
[9]	LUO X W, CHEN L, ZHOU H L, et al. A survey of underwater acoustic target recognition methods based on machine learning[J]. Journal of Marine Science and Engineering, 2023, 11(2):384.
[10]	DOAN S V, HUYNH-THE T, KIM D S. Underwater acoustic target classification based on dense convolutional neural network[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 99:1-5.
[11]	LI J F, ZHAO G F, LI B B, et al. A reduced dimension multiple signal classification-based direct location algorithm with dense arrays[J]. International Journal of Distributed Sensor Networks, 2022, 18(5):7153-7160.
[12]	TIAN S Z, BAI D, ZHOU J L, et al. Few-shot learning for joint model in underwater acoustic target recognition[J]. Scientific Reports, 2023, 13(1): 17502.
[13]	IRFAN M, JIANGBIN Z, ALI S, et al. DeepShip: an underwater acoustic benchmark dataset and a separable convolution based autoencoder for classification[J]. Expert Systems with Applications, 2021, 183(5):115270.
[14]	XIE Y, REN J W, XU J. Guiding the underwater acoustic target recognition with interpretable contrastive learning[C]// Proceedings of the OCEANS 2023.Limerick, Ireland: IEEE, 2023:1-6.
[15]	徐宝国, 何小杭, 魏智唯, 等. 基于运动想象脑电的机器人连续控制系统研究[J]. 仪器仪表学报, 2018, 39(9):10-19.
	XU B G, HE X H, WEI Z W, et al. Research on continuous control system for robot based on motor imagery EEG[J]. Chinese Journal of Scientific Instrument, 2018, 39(9):10-19. (in Chinese)
[16]	VAUGHAN T M, HEETDERKS W J, TREJO L J, et al. Brain-computer interface technology:a review of the Second International Meeting[J]. IEEE Transactions on Neural Systems & Rehabilitation Engineering A Publication of the IEEE Engineering in Medicine & Biology Society, 2003, 11(2):94-109.
[17]	NAKANISHI M, WANG Y J, CHEN X G, et al. Enhancing detection of SSVEPs for a high-speed brain speller using task-related component analysis[J]. IEEE Transactions on Biomedical Engineering, 2017, 65(1):104-112.
[18]	HERMAN P, PRASAD G, MCGINNITY T M, et al. Comparative analysis of spectral approaches to feature extraction for EEG-based motor imagery classification[J]. IEEE Transactions on Neural Systems & Rehabilitation Engineering, 2008, 16(4):317-326.
[19]	GRAMANN K, GWIN J T, BIGDELY-SHAMLO N, et al. Visual evoked responses during standing and walking[J]. Frontiers in Human Neuroscience, 2010, 4:202. doi: 10.3389/fnhum.2010.00202 pmid: 21267424
[20]	WANG R D, LIU Y, SHI J T, et al. Sound target detection under noisy environment using brain-computer interface[J]. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 2023, 31:229-237.
[21]	HAN X K, NIU J Y, GUO S J. A tactile-based brain computer interface P300 paradigm using vibration frequency and spatial location[J]. Journal of Medical and Biological Engineering, 2020, 40:773-782.
[22]	KOJIMA S, KANOH S I. An auditory brain-computer interface based on selective attention to multiple tone streams[J]. Plos One, 2024, 19(5):e0303565.
[23]	HUANG W C, ZHANG P Q, YU T Y, et al. A P300-based BCI system using stereoelectroencephalography and its application in a brain mechanistic study[J]. IEEE Transactions on Biomedical Engineering, 2020, 68(8):2509-2519.
[24]	FRISTON K J. Functional and effective connectivity in neuroimaging: a synthesis[J]. Human Brain Mapping, 1994, 2(1/2):56-78.
[25]	GRANGER C W. Investigating causal relations by econometric models and cross-spectral methods[J]. Econometrica: Journal of the Econometric Society, 1969, 37(3):424-438.
[26]	SETH A. Causalities connectivity of evolved neural networks during behavior[J]. Network:Computation in Neural Network Systems, 2005, 16(1):35-54.
[27]	SCHREIBER T. Measuring information transfer[J]. Physical Review Letters, 2000, 85(2):461-464. pmid: 10991308
[28]	PALVA J M, PALVA S, KAILA K. Phase synchrony among neuronal oscillations in the human cortex[J]. Journal of Neuroscience, 2005, 25(15):3962-3972. doi: 10.1523/JNEUROSCI.4250-04.2005 pmid: 15829648
[29]	YUAN J J, LUO Y J, YAN J H, et al. Neural correlates of the females’ susceptibility to negative emotions: an insight into gender-related prevalence of affective disturbances[J]. Human Brain Mapping, 2009, 30(11):3676-3686.
[30]	CHEN Y W, LIN C J. Combining SVMs with various feature selection strategies[J]. Studies in Fuzziness & Soft Computing, 2008, 207:315-324.
[31]	XIE J Y, WANG C X, JIANG S, et al. Feature selection method combing improved F-score and support vector machine[J]. Journal of Computer Applications, 2010, 30(4): 993-996.
[32]	DELORME A, MAKEIG S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis[J]. Journal of Neuroscience Methods, 2004, 134(1):9-21. doi: 10.1016/j.jneumeth.2003.10.009 pmid: 15102499
[33]	MOGNON A, JOVICICH J, BRUZZONE L, et al. ADJUST: an automatic EEG artifact detector based on the joint use of spatial and temporal features[J]. Psychophysiology, 2011, 48(2):229-240. doi: 10.1111/j.1469-8986.2010.01061.x pmid: 20636297
[34]	KARTON I, BACHMANN T. Disrupting dorsolateral prefrontal cortex by rTMS reduces the P300 based marker of deception[J]. Brain and Behavior, 2017, 7(4):e00656.
[35]	MOLHOLM S, SEHATPOUR P, MEHTA A D, et al. Audio-visual multisensory integration in superior parietal lobule revealed by human intracranial recordings[J]. Journal of Neurophysiology, 2006, 96(2):721-729. doi: 10.1152/jn.00285.2006 pmid: 16687619