基于改进BPNN的声速场分层建模

doi:10.11947/j.AGCS.2024.20230035

摘要/Abstract

摘要：

针对BPNN在三维声速场建模应用中存在的预测精度低、极易陷入局部最优及可解释性弱等局限性，提出了一种基于BPNN的温盐场建模方法，并联合声速经验公式设计了声速分层建模方案。同时通过引入自适应粒子群优化算法，对BPNN函数和架构进行改进和优化，提高了温盐场建模精度。采用中国南海中部区域的BOA_Argo网格数据对所提算法的建模性能进行试验，结果表明，本文算法能够充分反映海洋声速场物理特性，相较于传统算法其建模精度更高、稳健性更强，且具有优秀的稳定性和可靠性。

关键词: 三维声速场, 反向传播神经网络, 粒子群优化算法, BOA_Argo网格数据

Abstract:

Aiming at the limitations of BPNN in the application of 3D sound velocity field modeling, such as low prediction accuracy, easy to fall into the local optimum, and weak interpretability, a method of the thermohaline field modeling based on BPNNs is proposed, and a sound velocity hierarchical modeling scheme is designed jointly with the empirical equation of sound velocity. Meanwhile, the BPNN function and architecture are improved and optimized by introducing an adaptive particle swarm optimization algorithm to improve the accuracy of the thermohaline field modeling. Experiments on the modeling performance of the proposed algorithm are conducted using BOA_Argo grid data in the central region of the South China Sea. The results show that the algorithm proposed in this paper can fully reflect the physical properties of the ocean sound velocity field, with higher modeling accuracy and robustness than traditional algorithms, and with excellent stability and reliability.

Key words: 3D sound velocity field, back propagation neural network, particle swarm optimization algorithm, BOA_Argo grid data

中图分类号:

P229

王朝莹, 柴洪洲, 杜祯强. 基于改进BPNN的声速场分层建模[J]. 测绘学报, 2024, 53(12): 2316-2327.

Zhaoying WANG, Hongzhou CHAI, Zhenqiang DU. Hierarchical modeling of sound velocity field based on improved BPNN[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(12): 2316-2327.

图/表 13

图1

图2

图3

图4

图5

图6

图7

表1

表2

图8

图9

表3

表4

参考文献 34

[1]	杨元喜, 徐天河, 薛树强. 我国海洋大地测量基准与海洋导航技术研究进展与展望[J]. 测绘学报, 2017, 46(1): 1-8. DOI:. doi: 10.11947/j.AGCS.2017.20160519
	YANG Yuanxi, XU Tianhe, XUE Shuqiang. Progresses and prospects in developing marine geodetic datum and marine navigation of China[J]. Acta Geodaetica et Cartographica Sinica, 2017, 46(1): 1-8. DOI:. doi: 10.11947/j.AGCS.2017.20160519
[2]	辛明真, 阳凡林, 薛树强, 等. 顾及波束入射角的常梯度声线跟踪水下定位算法[J]. 测绘学报, 2020, 49(12): 1535-1542. DOI:. doi: 10.11947/j.AGCS.2020.20190518
	XIN Mingzhen, YANG Fanlin, XUE Shuqiang, et al. Constant-gradient acoustic line tracking underwater localization algorithm taking beam incidence angle into account[J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(12): 1535-1542. DOI:. doi: 10.11947/j.AGCS.2020.20190518
[3]	闫凤池, 王振杰, 赵爽, 等. 顾及双程声径的常梯度声线跟踪水下定位算法[J]. 测绘学报, 2022, 51(1): 31-40. DOI:. doi: 10.11947/j.AGCS.2022.20210234
	YAN Fengchi, WANG Zhenjie, ZHAO Shuang, et al. Constant gradient acoustic line tracking underwater localization algorithm considering two-way acoustic path[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(1): 31-40. DOI:. doi: 10.11947/j.AGCS.2022.20210234
[4]	张盛秋, 杨元喜, 徐天河. 基于GNSS-A的海洋声速变化估计及其对定位的影响[J]. 地球物理学报, 2023, 66(3): 961-972.
	ZHANG Shengqiu, YANG Yuanxi, XU Tianhe. Estimation of ocean sound speed variation based on GNSS-A and its effect on localization[J]. Chinese Journal of Geophysics, 2023, 66(3): 961-972.
[5]	徐博. 水下多AUV协同定位方法[M]. 北京: 国防工业出版社, 2019.
	XU Bo. Cooperative localization for multiple autonomous underwater vehicles[M]. Beijing: National Defense Industry Press, 2019.
[6]	刘伯胜. 水声学原理[M]. 3版. 北京: 科学出版社, 2019.
	LIU Bosheng. Principles of underwater acoustics[M]. 3rd ed. Beijing: Science Press, 2019.
[7]	ZHANG L H, LIU X P, JIA S D, et al. A line-surface integrated algorithm for underwater terrain matching[J]. The Journal of Geodesy and Geoinformation Science, 2019, 2(4): 10-20.
[8]	MUNK W H. Sound channel in an exponentially stratified ocean, with application to SOFAR[J]. The Journal of the Acoustical Society of America, 1974, 55(2): 220-226.
[9]	DAVIS T, COUNTRYMAN K, CARRON M. Tailored acoustic products utilizing the navoceano gdem (a generalized digital environmental model)[C]//Proceedings of the 36th Naval Symposium on Underwater Acoustics. San Diego: Naval Ocean Systems Center, 1986.
[10]	张旭, 张永刚, 张健雪, 等. 一种新的声速剖面结构参数化方法[J]. 海洋学报(中文版), 2011, 33(5): 54-60.
	ZHANG Xu, ZHANG Yonggang, ZHANG Jianxue, et al. A new model for calculating sound speed profile structure[J]. Haiyang Xuebao, 2011, 33(5): 54-60.
[11]	NISTAD J G, WESTFELD P. Improved techniques to resolve the water column sound speed structure for multibeam ray-tracing[J]. The International Hydrographic Review, 2022, 27: 35-54.
[12]	LORENZ E N. Empirical orthogonal functions and statistical weather prediction[J]. Scientific Reports, 1956, 409(2): 997-999.
[13]	李洪超, 文汉江, 蔡艳辉, 等. 基于Argo浮标和EOF建立区域海水三维声速场的方法[J]. 测绘科学, 2012, 37(2): 74-76.
	LI Hongchao, WEN Hanjiang, CAI Yanhui, et al. Modeling three-dimensional acoustic field in the ocean by using Argo and EOF[J]. Science of Surveying and Mapping, 2012, 37(2): 74-76.
[14]	OU Zhenyi, QU Ke, LIU Chen. Estimation of sound speed profiles using a random forest model with satellite surface observations[J]. Shock and Vibration, 2022, 2022: 2653791.
[15]	YU Xiaokang, XU Tianhe, WANG Junting. Sound velocity profile prediction method based on RBF neural network[C]//Proceedings of 2020 China Satellite Navigation Conference. Beijing: Springer, 2020: 475-487.
[16]	LI Bingyang, ZHAI Jingsheng. A novel sound speed profile prediction method based on the convolutional long-short term memory network[J]. Journal of Marine Science and Engineering, 2022, 10(5): 572.
[17]	LEBLANC L R, MIDDLETON F H. An underwater acoustic sound velocity data model[J]. The Journal of the Acoustical Society of America, 1980, 67(12): 2055-2062.
[18]	DAVIS R E. Predictability of sea surface temperature and sea level pressure anomalies over the North Pacific Ocean[J]. Journal of Physical Oceanography, 1976, 6(3): 249-266.
[19]	TOLSTOY A, DIACHOK O, FRAZER L N. Acoustic tomography via matched field processing[J]. The Journal of the Acoustical Society of America, 1991, 89(3): 1119-1127.
[20]	NAJAH A, EL-SHAFIE A, KARIM O A, et al. Application of artificial neural networks for water quality prediction[J]. Neural Computing and Applications, 2013, 22(S1): 187-201.
[21]	孙行, 黄泽纯. 面向复杂地形区气温场拟合的回归学习方法精度比较[J]. 测绘与空间地理信息, 2022, 45(5): 18-23.
	SUN Xing, HUANG Zechun. Accuracy comparison of regression learning methods for fitting temperature fields in complex terrain areas[J]. Mapping and Spatial Geographic Information, 2022, 45(5): 18-23.
[22]	WANG Junting, XU Tianhe, NIE Wenfeng, et al. The construction of sound speed field based on back propagation neural network in the global ocean[J]. Marine Geodesy, 2020, 43(6): 621-642.
[23]	孙佳龙, 张杰, 唐玥, 等. 双种群约束QPSO-BP的声速剖面反演方法[J]. 测绘科学, 2021, 46(8): 127-134.
	SUN Jialong, ZHANG Jie, TANG Yue, et al. Inversion method of sound velocity profile based on QPSO-BP with two population constraints[J]. Science of Surveying and Mapping, 2021, 46(8): 127-134.
[24]	胡合欢, 崔永胜, 周进忠, 等. BP神经网络在构建声速场中的应用研究[J]. 海洋测绘, 2015, 35(5): 67-70.
	HU Hehuan, CUI Yongsheng, ZHOU Jinzhong, et al. Application of back-propagation neural network in establishing velocity fields[J]. Hydrographic Surveying and Charting, 2015, 35(5): 67-70.
[25]	李林洋, 徐天河, 王君婷, 等. 联合匹配场和神经网络的声速时间场构建方法[J]. 哈尔滨工程大学学报, 2023, 44(11): 2044-2053.
	LI Linyang, XU Tianhe, WANG Junting, et al. A method for constructing sound velocity time field by combining matched field and neural network[J]. Journal of Harbin Engineering University, 2023, 44(11): 2044-2053.
[26]	HUANG Jin, LUO Yu, SHI Jian, et al. Rapid modeling of the sound speed field in the South China Sea based on a comprehensive optimal LM-BP artificial neural network[J]. Journal of Marine Science and Engineering, 2021, 9(5): 488.
[27]	周梦, 吕志刚, 邸若海, 等. 基于小样本数据的BP神经网络建模[J]. 科学技术与工程, 2022, 22(7): 2754-2760.
	ZHOU Meng, LÜ Zhigang, DI Ruohai, et al. BP neural network modeling based on small sample data[J]. Science Technology and Engineering, 2022, 22(7): 2754-2760.
[28]	钱建国, 樊意广. 基于改进小波神经网络的GPS高程拟合研究[J]. 大地测量与地球动力学, 2022, 42(3): 253-257.
	QIAN Jianguo, FAN Yiguang. Research on GPS height fitting based on improved wavelet neural network[J]. Journal of Geodesy and Geodynamics, 2022, 42(3): 253-257.
[29]	RUMELHART D E, HINTON G E, WILLIAMS R J. Learning representations by back-propagating errors[J]. Nature, 1986, 323: 533-536.
[30]	DEMUTH H B, BEALE M H, JESS O D, HAGAN M T. Neural network design[M]. 2nd ed. Oklahoma: Martin Hagan, 2014.
[31]	周丰年, 赵建虎, 周才扬. 多波束测深系统最优声速公式的确定[J]. 台湾海峡, 2001(4): 411-419.
	ZHOU Fengnian, ZHAO Jianhu, ZHOU Caiyang. Determination of the optimal sound velocity equation for multibeam bathymetric systems[J]. Taiwan Strait, 2001(4): 411-419.
[32]	SHI Y, EBERHART R. A modified particle swarm optimizer[C]//Proceedings of 1998 International Conference on Evolutionary Computation. Anchorage: IEEE, 1998: 69-73.
[33]	黄洋, 鲁海燕, 许凯波, 等. 基于S型函数的自适应粒子群优化算法[J]. 计算机科学, 2019, 46(1): 245-250.
	HUANG Yang, LU Haiyan, XU Kaibo, et al. S-shaped function based adaptive particle swarm optimization algorithm[J]. Computer Science, 2019, 46(1): 245-250.
[34]	LI Hong, XU Fanghua, ZHOU Wei, et al. Development of a global gridded Argo data set with barnes successive corrections: a new global gridded Argo data set[J]. Journal of Geophysical Research: Oceans, 2017, 122(2): 866-889.

算法类型	声速误差/（m²/s²）
EOF	0.461 016
IBPNN_c	0.251 132
IBPNN	0.032 693
PSO-IBPNN	0.022 245
APSO-IBPN N	0.024 198

算法类型	建模时间/s
EOF	0.218
IBPNN_c	11
IBPNN	19
PSO-IBPNN	4786
APSO-IBPNN	2458

算法简称	No.1	No.12	No.23	No.43	No.48
PSO-IBPNN	0.258 734	0.044 405	0.213 312	0.035 571	0.061 199
APSO-IBPNN	0.064 499	0.052 736	0.114 254	0.043 104	0.064 571

算法类型	预测耗时
EOF	0.005 8
IBPNN_c	0.000 1
IBPNN	0.266 6
PSO-IBPN N	0.173 0
APSO-IBPNN	0.250 6