利用GA-NN模型反演声速剖面的众源水深数据声速改正

袁浩, 贾帅东, 金绍华, 张立华, 王华

袁浩, 贾帅东, 金绍华, 张立华, 王华. 利用GA-NN模型反演声速剖面的众源水深数据声速改正[J]. 武汉大学学报 ( 信息科学版), 2023, 48(3): 377-385. DOI: 10.13203/j.whugis20200515
引用本文: 袁浩, 贾帅东, 金绍华, 张立华, 王华. 利用GA-NN模型反演声速剖面的众源水深数据声速改正[J]. 武汉大学学报 ( 信息科学版), 2023, 48(3): 377-385. DOI: 10.13203/j.whugis20200515
YUAN Hao, JIA Shuaidong, JIN Shaohua, ZHANG Lihua, WANG Hua. Correction for Crowd Sourced Bathymetry Data Using GA-NN Model to Inverse Sound Velocity Profiles[J]. Geomatics and Information Science of Wuhan University, 2023, 48(3): 377-385. DOI: 10.13203/j.whugis20200515
Citation: YUAN Hao, JIA Shuaidong, JIN Shaohua, ZHANG Lihua, WANG Hua. Correction for Crowd Sourced Bathymetry Data Using GA-NN Model to Inverse Sound Velocity Profiles[J]. Geomatics and Information Science of Wuhan University, 2023, 48(3): 377-385. DOI: 10.13203/j.whugis20200515

利用GA-NN模型反演声速剖面的众源水深数据声速改正

基金项目: 

国家自然科学基金 41901320

国家自然科学基金 41871369

国家自然科学基金 41774014

详细信息
    作者简介:

    袁浩,硕士,主要从事众源测深数据处理研究。573008183@qq.com

    通讯作者:

    贾帅东,博士,讲师。sky_jsd@163.com

  • 中图分类号: P237

Correction for Crowd Sourced Bathymetry Data Using GA-NN Model to Inverse Sound Velocity Profiles

  • 摘要: 针对当前众源水深数据后处理过程中缺少高精度的实测声速剖面,导致测深数据质量偏低的现状,提出了一种基于遗传算法优化反向传播神经网络(genetic algorithm-back propagation neural network,GA-NN)模型反演声速剖面的声速改正方法。首先,利用历史声速剖面群进行正交经验函数分析,提取特征向量与重构系数范围;然后,结合海区的历史声速场数据训练GA-NN模型;最后,将海表声速数据输入模型反演声速剖面,并分析不同方法下的声速剖面分别进行声速改正后的水深和位置误差。实验结果表明,在复杂的海底地形下,与现有方法相比,所提方法反演的声速剖面更适用于众源水深数据的声速改正,削弱了声速误差的影响,提高了众源水深数据的处理精度。
    Abstract:
      Objectives  Currently , the measured sound velocity profile with high accuracy is lack in the processing of the crowd sourced bathymetry(CSB) data, so that the quality of the sounding data is low. Aiming at this situation above, a method for utilizing the inversion sound velocity profile obtained from the BP neural network model optimized by the genetic algorithm (GA-NN model) is proposed to correct the CSB data.
      Methods  Firstly, the eigenvector and the reconstruction coefficient range are extracted from the results via empirical orthogonal functions analysis of the historical sound velocity profile group. Secondly, GA-NN model is trained by utilizing the information of the historical sound velocity profile field. Finally, the surface sound velocity is inputted into the model for getting the inversion sound velocity profile, and statistically analyzed the depth and position errors after correcting the data via different sound velocity profiles.
      Results  The experimental results show that in the sea area with complex seafloor topography, the error index of sound velocity profile inverted by the proposed method is smaller, the corrected seabed topography is more fitting to the simulated seabed, and the error of water depth and position is smaller.
      Conclusions  The sound velocity profile inverted by the proposed method is more suitable for the sound speed correction in the CSB work, weaks the influence of sound velocity error on seafloor topography, and improves the precision of CSB data.
  • 图  1   声速剖面反演及改正流程图

    Figure  1.   Flowchart of Inversion and Correction of SVP

    图  2   BP神经网络结构

    Figure  2.   Architecture of BP Neural Network

    图  3   声速剖面分布图

    Figure  3.   Distribution of Sound Velocity Profiles

    图  4   1号海域的声速剖面对比图

    Figure  4.   Comparison of Sound Velocity Profiles in Experimental Area 1

    图  5   声速改正后海底地形图(1号海域)

    Figure  5.   Seafloor Corrected by Sound Velocity Profiles(Area 1)

    图  6   2号海域的声速剖面对比图

    Figure  6.   Comparison of Sound Velocity Profiles in Experimental Area 2

    图  7   声速改正后海底地形图(2号海域)

    Figure  7.   Seafloor Corrected by Sound Velocity Profiles(Area 2)

    表  1   前7阶重构系数搜索范围

    Table  1   Seven Orders Reconstruction Coefficients Search Ranges

    实验海区 搜索范围 重构系数
    $ {a}_{1} $ $ {a}_{2} $ $ {a}_{3} $ $ {a}_{4} $ $ {a}_{5} $ $ {a}_{6} $ $ {a}_{7} $
    1号海域 上限 34.105 42.019 23.930 9.228 7.752 3.616 8.175
    下限 -57.090 -15.239 -20.078 -9.713 -7.117 -6.480 -7.691
    2号海域 上限 45.724 30.264 16.939 8.064 4.038 3.021 -1.217
    下限 -80.052 -16.536 -10.272 -7.108 -3.927 -3.049 -1.656
    下载: 导出CSV

    表  2   声速剖面误差统计结果(1号海域)/$ (\mathrm{m}\bullet {\mathrm{s}}^{-1}) $

    Table  2   Statistics of Sound Velocity Profile Errors(Area 1)/$ (\mathrm{m}\bullet {\mathrm{s}}^{-1}) $

    改正方法 最大值 最小值 平均值 中误差
    方法1 9.8 1.1 4.6 1.7
    方法2 4.7 -0.5 1.7 1.3
    方法3 7.9 -1.4 2.9 2.4
    方法4 海表声速源 4.4 -2.7 0.5 1.6
    遥感数据源 4.6 -2.6 0.7 1.6
    下载: 导出CSV

    表  3   水深误差统计结果(1号海域)/m

    Table  3   Statistics of Depth Errors(Area 1)/m

    改正方法 最大值 最小值 平均值 中误差
    方法1 9.389 8 -0.826 3 0.084 2 2.056 6
    方法2 8.696 2 -0.296 6 0.662 7 1.881 7
    方法3 12.060 3 0.497 2 1.948 5 2.494 2
    方法4 海表声速源 2.665 8 -0.097 6 0.210 6 0.586 6
    遥感数据源 3.101 5 -0.129 4 0.222 5 0.681 9
    下载: 导出CSV

    表  4   位置误差统计结果(1号海域)/m

    Table  4   Statistics of Position Errors(Area 1)/m

    改正方法 最大值 最小值 平均值 中误差
    方法1 26.081 6 -2.295 3 0.233 8 5.712 6
    方法2 23.048 1 -0.980 1 1.572 0 5.041 6
    方法3 34.074 6 1.404 8 5.502 3 7.047 0
    方法4 海表声速源 7.386 3 -0.270 5 0.583 6 1.625 2
    遥感数据源 8.588 4 -0.358 2 0.616 1 1.888 2
    下载: 导出CSV

    表  5   声速剖面误差统计结果(2号海域)/$ (\mathrm{m}\bullet {\mathrm{s}}^{-1}) $

    Table  5   Statistics of Sound Velocity Profile Errors(Area 2)/$ (\mathrm{m}\bullet {\mathrm{s}}^{-1}) $

    改正方法 最大值 最小值 平均值 中误差
    方法1 4.5 -1.2 0.1 1.0
    方法2 2.7 -2.0 -0.6 0.7
    方法3 7.8 -2.0 1.3 2.0
    方法4 海表声速源 2.6 -1.3 -0.2 0.8
    遥感数据源 2.7 -1.7 -0.2 0.9
    下载: 导出CSV

    表  6   水深误差统计结果(2号海域)/m

    Table  6   Statistics of Depth Errors(Area 2)/m

    改正方法 最大值 最小值 平均值 中误差
    方法1 2.554 0 0.006 7 0.392 2 0.610 0
    方法2 0.798 5 0.110 5 0.178 6 0.128 1
    方法3 9.589 8 -0.241 2 0.919 8 2.124 9
    方法4 海表声速源 0.208 4 -0.100 5 0.006 5 0.052 1
    遥感数据源 0.662 0 0.026 1 0.065 0 0.107 3
    下载: 导出CSV

    表  7   位置误差统计结果(2号海域)/m

    Table  7   Statistics of Position Errors(Area 2)/m

    改正方法 最大值 最小值 平均值 中误差
    方法1 6.862 5 0.018 0 1.053 7 1.639 1
    方法2 2.136 2 0.295 7 0.477 9 0.342 7
    方法3 25.739 4 -0.647 4 2.468 8 5.703 3
    方法4 海表声速源 0.558 4 -0.269 4 0.017 3 0.139 6
    遥感数据源 1.772 1 0.069 8 0.174 0 0.287 4
    下载: 导出CSV
  • [1]

    International Hydrographic Organization. B-12 Edition 2.0.3, Guidance on Crowdsourced Bathymetry[EB/OL]. (2020-01-20)[2021-04-29]. https://iho.int/uploads/user/pubs/bathy/B_12_Ed2.0.3_2020.pdf.

    [2]

    Thornton T. TeamSurv-Surveying with the Crowd: The Application of Crowdsourcing to Hydrographic Surveying[EB/OL]. (2011-07-15)[2020-12-23]. https://www.hydro-international.com/content/article/teamsurv-surveying-with-the-crowd.

    [3]

    Calder B R, Dijkstra S J, Hoy S, et al. A Design for a Trusted Community Bathymetry System[J]. Marine Geodesy, 2020, 43(4): 327-358. doi: 10.1080/01490419.2020.1718255

    [4]

    Arfeen K. Automatic Processing of Arctic Crowd-Sourced Hydrographic Data While Improving Bathymetric Accuracy and Uncertainty Assessment[D]. Canada Fredericton: University of New Brunswick, 2019.

    [5] 黄辰虎, 王梓, 张晓娟, 等. 联合WOA2018温盐模型及实测温盐资料重构全深度声速剖面(二): 精度评估[J]. 海洋测绘, 2020, 40(2): 9-14. doi: 10.3969/j.issn.1671-3044.2020.02.002

    Huang Chenhu, Wang Zi, Zhang Xiaojuan, et al. Reconstruction of the Full-Depth Sound Speed Profile with World Ocean Atlas 2018 and In-situ Temperature and Salinity in Sea Water, PartⅡ: Accuracy Assessment[J]. Hydrographic Surveying and Charting, 2020, 40(2): 9-14. doi: 10.3969/j.issn.1671-3044.2020.02.002

    [6] 黄辰虎, 陆秀平, 王克平, 等. 联合XBT和WOA13模型盐度信息的深水走航声速准确确定[J]. 海洋通报, 2016, 35(5): 554-561.

    Huang Chenhu, Lu Xiuping, Wang Keping, et al. A Method of Exactly Determining the Sound Velocity in Deep Water Based on Salt Information from WOA13 Model and XBT Data[J]. Marine Science Bulletin, 2016, 35(5): 554-561.

    [7] 黄辰虎, 陆秀平, 叶安娜, 等. 海底地形测量成果的质量检核评估(二): 深远海海域声速剖面的获取[J]. 海洋测绘, 2017, 37(3): 12-16. doi: 10.3969/j.issn.1671-3044.2017.03.002

    Huang Chenhu, Lu Xiuping, Ye Anna, et al. Examination and Assessment of Bathymetric Surveying Products, Part Ⅱ: Acquisition of Sound Speed Profiles in the Deep Sea[J]. Hydrographic Surveying and Charting, 2017, 37(3): 12-16. doi: 10.3969/j.issn.1671-3044.2017.03.002

    [8] 赵建虎, 刘经南. 多波束测深及图像数据处理[M]. 武汉: 武汉大学出版社, 2008.

    Zhao Jianhu, Liu Jingnan. Multi-beam Sounding and Image Data Processing[M]. Wuhan: Wuhan University Press, 2008.

    [9] 孙文川, 暴景阳, 金绍华, 等. 多波束海底地形畸变校正与声速剖面反演[J]. 武汉大学学报(信息科学版), 2016, 41(3): 349-355. doi: 10.13203/j.whugis20140142

    Sun Wenchuan, Bao Jingyang, Jin Shaohua, et al. Inversion of Sound Velocity Profiles by Correcting the Terrain Distortion[J]. Geomatics and Information Science of Wuhan University, 2016, 41(3): 349-355. doi: 10.13203/j.whugis20140142

    [10] 张志伟, 暴景阳, 肖付民, 等. 利用模拟退火算法反演多波束测量声速剖面[J]. 武汉大学学报(信息科学版), 2018, 43(8): 1234-1241. doi: 10.13203/j.whugis20160304

    Zhang Zhiwei, Bao Jingyang, Xiao Fumin, et al. Inversion of Sound Velocity Profile in Multibeam Survey Based on Simulated Annealing Algorithm[J]. Geomatics and Information Science of Wuhan University, 2018, 43(8): 1234-1241. doi: 10.13203/j.whugis20160304

    [11] 李梦竹, 李整林, 李倩倩. 南海北部负跃层环境下海底参数声学反演[J]. 声学学报, 2019, 44(3): 321-328.

    Li Mengzhu, Li Zhenglin, Li Qianqian. Geoacoustic Inversion for Bottom Parameters in a Thermocline Environment in the Northern Area of the South China Sea[J]. Acta Acustica, 2019, 44(3): 321-328.

    [12] 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. doi: 10.3969/j.issn.1671-3044.2015.05.017

    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. doi: 10.3969/j.issn.1671-3044.2015.05.017

    [13] 沈远海, 马远良, 屠庆平, 等. 浅水声速剖面的反演方法与实验验证[J]. 西北工业大学学报, 2000, 18(2): 212-215. doi: 10.3969/j.issn.1000-2758.2000.02.010

    Shen Yuanhai, Ma Yuanliang, Tu Qingping, et al. Inversion of Sound Speed Profile for Shallow-Water Environment with Experimental Verification[J]. Journal of Northwestern Polytechnical University, 2000, 18(2): 212-215. doi: 10.3969/j.issn.1000-2758.2000.02.010

    [14] 唐俊峰. 海水中的声速剖面反演研究[D]. 哈尔滨: 哈尔滨工程大学, 2006.

    Tang Junfeng. Research on Inversion of Sound Speed Profile in the Ocean[D]. Harbin: Harbin Engineering University, 2006.

    [15]

    US Department of Commerce, National Oceanic and Atmospheric Administration. NOS Hydrographic Surveys Specifications and Deliverables[EB/OL]. (2020-01-20)[2021-04-29]. https://nauticalcharts.noaa.gov/publications/docs/standards-and-requirements/specs/hssd-2016.pdf.

    [16] 李家彪. 多波束勘测原理技术与方法[M]. 北京: 海洋出版社, 1999.

    Li Jiabiao. Principle, Technology and Method of Multi-beam Survey[M]. Beijing: Ocean Press, 1999.

    [17] 艾锐峰, 程杰, 欧阳军, 等. 海区声速剖面的现场反演方法[J]. 计算机应用, 2015, 35(S1): 327-330.

    Ai Ruifeng, Cheng Jie, Ouyang Jun, et al. On-Line Retrieval Methodology for Sound Speed Profile of Sea Area[J]. Journal of Computer Applications, 2015, 35(S1): 327-330.

    [18] 丁继胜, 周兴华, 唐秋华, 等. 多波束勘测声速剖面场的EOF表示方法[J]. 武汉大学学报(信息科学版), 2007, 32(5): 446-449. http://ch.whu.edu.cn/article/id/1892

    Ding Jisheng, Zhou Xinghua, Tang Qiuhua, et al. Expression of Multi-beam Echo Sounding Sound Velocity Profile with Empirical Orthogonal Functions[J]. Geomatics and Information Science of Wuhan University, 2007, 32(5): 446-449. http://ch.whu.edu.cn/article/id/1892

    [19] 陆秀平, 边少锋, 黄谟涛, 等. 常梯度声线跟踪中平均声速的改进算法[J]. 武汉大学学报(信息科学版), 2012, 37(5): 590-593. http://ch.whu.edu.cn/article/id/209

    Lu Xiuping, Bian Shaofeng, Huang Motao, et al. An Improved Method for Calculating Average Sound Speed in Constant Gradient Sound Ray Tracing Technology[J]. Geomatics and Information Science of Wuhan University, 2012, 37(5): 590-593. http://ch.whu.edu.cn/article/id/209

    [20] 赵祥鸿, 暴景阳, 欧阳永忠, 等. 利用BP神经网络剔除多波束测深数据粗差[J]. 武汉大学学报(信息科学版), 2019, 44(4): 518-524. doi: 10.13203/j.whugis20160336

    Zhao Xianghong, Bao Jingyang, Ouyang Yongzhong, et al. Detecting Outlier of Multibeam Sounding with BP Neural Network[J]. Geomatics and Information Science of Wuhan University, 2019, 44(4): 518-524. doi: 10.13203/j.whugis20160336

    [21] 尹淑玲, 舒宁, 刘新华. 基于自适应遗传算法和改进BP算法的遥感影像分类[J]. 武汉大学学报(信息科学版), 2007, 32(3): 201-204. http://ch.whu.edu.cn/article/id/1835

    Yin Shuling, Shu Ning, Liu Xinhua. Classification of Remote Sensing Image Based on Adaptive GA and Improved BP Algorithm[J]. Geomatics and Information Science of Wuhan University, 2007, 32(3): 201-204. http://ch.whu.edu.cn/article/id/1835

    [22] 陈鹏. 遗传算法的改进及在调度优化中的应用研究[D]. 武汉: 中国地质大学, 2011.

    Chen Peng. Improvement of Genetic Algorithm and Using in Application and Research of Scheduling Optimization[D]. Wuhan: China University of Geoscien‑ ces, 2011.

图(7)  /  表(7)
计量
  • 文章访问数:  470
  • HTML全文浏览量:  100
  • PDF下载量:  61
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-09-27
  • 网络出版日期:  2023-03-23
  • 发布日期:  2023-03-04

目录

    /

    返回文章
    返回