留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于多浮标空间多特征融合的海水溶解氧浓度预测

朱奇光 申震 李享 魏祯 乔文静 张淋淞 陈颖

朱奇光,申震,李享,等. 基于多浮标空间多特征融合的海水溶解氧浓度预测[J]. 海洋学报,2025,47(x):1–13
引用本文: 朱奇光,申震,李享,等. 基于多浮标空间多特征融合的海水溶解氧浓度预测[J]. 海洋学报,2025,47(x):1–13
Zhu Qiguang,Shen Zhen,Li Xiang, et al. Prediction of seawater dissolved oxygen concentration based on multi-buoy spatial multi-feature fusion[J]. Haiyang Xuebao,2025, 47(x):1–13
Citation: Zhu Qiguang,Shen Zhen,Li Xiang, et al. Prediction of seawater dissolved oxygen concentration based on multi-buoy spatial multi-feature fusion[J]. Haiyang Xuebao,2025, 47(x):1–13

基于多浮标空间多特征融合的海水溶解氧浓度预测

基金项目: 国家自然科学基金项目(62275228);河北省重点研发计划项目(19273901D;20373301D);河北省自然科学基金项目(D2024203002)。
详细信息
    作者简介:

    朱奇光(1978—),男,主要从事多传感器信息融合与大数据分析方面的研究。E-mail:zhu7880@ysu.edu.cn

    通讯作者:

    陈颖(1980—),女,主要从事海洋生态环境监测与大数据分析的研究。E-mail:chenying@ysu.edu.cn

  • 中图分类号: X834

Prediction of seawater dissolved oxygen concentration based on multi-buoy spatial multi-feature fusion

  • 摘要: 溶解氧浓度是衡量海水水质的重要指标之一。为了及时掌握海水水质变化情况,降低海水污染风险及其带来的损失,建立海洋水质参数预测机制至关重要。为此,本文提出了一种基于浮标网络时空信息融合和改进生成对抗网络(Generative Adversarial Networks, GAN)的海水溶解氧浓度预测模型,旨在整合监测区域内浮标网络的拓扑信息并实现浮标传感器的多特征融合。该模型利用图注意力网络(Graph Attention Mechanism,GAT)挖掘不同近邻点对目标节点的影响,计算邻接节点的权重,从而捕获浮标数据的时空特征;通过双头注意力机制与双时间尺度更新规则(Two Time-Scale Update Rule, TTUR)优化GAN预测网络及网络训练过程,改善生成对抗网络的训练速度平衡问题,提高生成器网络的拟合效果。以均方误差、均方根误差、平均绝对误差与决定系数为评价指标进行模型预测性能对比,结果表明,所提出模型的各项评价指标均优于其他模型,能够有效挖掘多浮标的空间信息,克服了传统方法在海水溶解氧浓度预测中存在的精度低、无法灵活利用历史空间数据、训练稳定性差和速度慢等不足,可为海洋水质监测及预测提供重要的技术支撑。
  • 图  1  浮标地理位置

    Fig.  1  Geographic location of buoy

    图  2  各浮标测得的溶解氧数据

    Fig.  2  Dissolved oxygen data measured by each buoy

    图  3  各浮标参数间相关性分析结果

    Fig.  3  Correlation analysis results between parameters of various buoys

    图  4  各浮标特征融合后的特征向量

    Fig.  4  The feature vector obtained by fusing the features of each buoy

    图  5  浮标间的拓扑关系

    Fig.  5  Topological relationship between buoys

    图  6  GAT模块结构

    Fig.  6  GAT module structure

    图  7  经GAT自适应空间特征提取后的邻接矩阵

    Fig.  7  Adjacency matrix after GAT adaptive spatial feature extraction

    图  8  SA-WGAN_GP模型

    Fig.  8  SA-WGAN-GP model

    图  9  不同预测模型的预测性能

    Fig.  9  The predictive performance of different prediction models

    表  1  浮标搭载传感器监测参数

    Tab.  1  Monitoring parameters of buoys equipped with sensors

    参数英文符号单位
    温度Temp
    电导率CondmS/cm
    盐度SalPpt
    溶解氧饱和度DOmg/L
    酸碱度pH1
    浊度TurbNTU
    叶绿素Chlug/L
    藻红蛋白PEug/L
    下载: 导出CSV

    表  2  不同浮标上所选择的特征

    Tab.  2  Features selected on different buoys

    浮标参数1参数2参数3参数4参数5
    GX-02ChlpHTempCondSal
    GX-14pHTempChlCondSal
    GX-15TempChlPECondSal
    GX-17TempCondpHSalTurb
    GX-03pHCondChlTempPE
    GX-19TempCondpHSalPE
    下载: 导出CSV

    表  3  浮标间的距离关系 (距离/km)

    Tab.  3  Distance relationship between buoys (distance/km)

    浮标GX-02GX-14GX-15GX-17GX-03GX-19
    GX-0203.3451.8059.2261.1671.48
    GX-143.34048.7356.2258.1068.36
    GX-1551.8048.7307.989.3919.80
    GX-1759.2256.227.9803.0313.61
    GX-0361.1658.109.393.03010.88
    GX-1971.4868.3619.8013.6110.880
    下载: 导出CSV

    表  4  不同模型的评价指标对比

    Tab.  4  Comparison of evaluation indicators for different models

    模型 MSE RMSE MAE $ {{\mathrm{R}}}^{2} $
    CNN-LSTM 0.596 0.772 0.502 0.138
    CNN-GRU 0.407 0.637 0.292 0.312
    GCN-WGAN_GP 0.313 0.559 0.278 0.639
    GAT-WGAN_GP 0.198 0.444 0.124 0.804
    下载: 导出CSV
  • [1] 齐庆华, 蔡榕硕, 颜秀花. 气候变化与我国海洋灾害风险治理探讨[J]. 海洋通报, 2019, 38(4): 361−367.

    Qi Qinghua, Cai Rongshuo, Yan Xiuhua. Discussion on climate change and marine disaster risk governance in the coastal China seas[J]. Marine Science Bulletin, 2019, 38(4): 361−367.
    [2] Kisi O, Alizamir M, Gorgij A D. Dissolved oxygen prediction using a new ensemble method[J]. Environmental Science and Pollution Research, 2020, 27(9): 9589−9603. doi: 10.1007/s11356-019-07574-w
    [3] Li Yanni, Yan Shiqiang, Shi Hongda, et al. Hydrodynamic analysis of a novel multi-buoy wind-wave energy system[J]. Renewable Energy, 2023, 219: 119477. doi: 10.1016/j.renene.2023.119477
    [4] Liu Shixuan, Song Miaomiao, Chen Shizhe, et al. An intelligent modeling framework to optimize the spatial layout of ocean moored buoy observing networks[J]. Frontiers in Marine Science, 2023, 10: 1134418. doi: 10.3389/fmars.2023.1134418
    [5] Jin Junchen, Rong Dingding, Zhang Tong, et al. A GAN-based short-term link traffic prediction approach for urban road networks under a parallel learning framework[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(9): 16185−16196. doi: 10.1109/TITS.2022.3148358
    [6] 杜先君, 李河. 基于集成GCN-Transformer网络的ENSO预测模型[J]. 海洋学报, 2023, 45(12): 156−165.

    Du Xianjun, Li He. ENSO prediction model based on integrated GCN-Transformer network[J]. Haiyang Xuebao, 2023, 45(12): 156−165.
    [7] Velickovic P, Cucurull G, Casanova A, et al. Graph attention networks[J]. arXiv: 1710.10903, 2017. (查阅网上资料, 不确定本条文献类型及格式是否正确, 请确认)
    [8] Li Wenjing, Fang Huaiyang, Qin Guangxiong, et al. Concentration estimation of dissolved oxygen in Pearl River Basin using input variable selection and machine learning techniques[J]. Science of the Total Environment, 2020, 731: 139099. doi: 10.1016/j.scitotenv.2020.139099
    [9] Wu Junhao, Wang Zhaocai, Dong Jinghan, et al. Multi-step ahead dissolved oxygen concentration prediction based on knowledge guided ensemble learning and explainable artificial intelligence[J]. Journal of Hydrology, 2024, 636: 131297. doi: 10.1016/j.jhydrol.2024.131297
    [10] Liu Yeqi, Gong Chuanyang, Yang Ling, et al. DSTP-RNN: a dual-stage two-phase attention-based recurrent neural network for long-term and multivariate time series prediction[J]. Expert Systems with Applications, 2020, 143: 113082. doi: 10.1016/j.eswa.2019.113082
    [11] Yue Weihao, Xu Yongsheng, Xiang Liang, et al. Prediction of 3-D ocean temperature based on self-attention and predictive RNN[J]. IEEE Geoscience and Remote Sensing Letters, 2024, 21: 1501305.
    [12] 石绥祥, 王蕾, 余璇, 等. 长短期记忆神经网络在叶绿素a浓度预测中的应用[J]. 海洋学报, 2020, 42(2): 134−142. doi: 10.3969/j.issn.0253-4193.2020.02.014

    Shi Suixiang, Wang Lei, Yu Xuan, et al. Application of long term and short term memory neural network in prediction of chlorophyll a concentration[J]. Haiyang Xuebao, 2020, 42(2): 134−142. doi: 10.3969/j.issn.0253-4193.2020.02.014
    [13] Jia Xiaoyan, Ji Qiyan, Han Lei, et al. Prediction of sea surface temperature in the East China Sea based on LSTM neural network[J]. Remote Sensing, 2022, 14(14): 3300. doi: 10.3390/rs14143300
    [14] 张宇, 许大志, 俞胜宾, 等. 基于多尺度深度学习对南海海表温度预报的研究[J]. 海洋学报, 2024, 46(5): 27−36. doi: 10.12284/hyxb2024034

    Zhang Yu, Xu Dazhi, Yu Shengbin, et al. Forecast of sea surface temperature in the South China Sea based on multi-scale deep learning model[J]. Haiyang Xuebao, 2024, 46(5): 27−36. doi: 10.12284/hyxb2024034
    [15] Goodfellow I J, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal: MIT Press, 2014: 2672−2680.
    [16] Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial networks[J]. Communications of the ACM, 2020, 63(11): 139−144. doi: 10.1145/3422622
    [17] Deng Jiaxiang, Deng Yong, Cheong K H. Combining conflicting evidence based on Pearson correlation coefficient and weighted graph[J]. International Journal of Intelligent Systems, 2021, 36(12): 7443−7460. doi: 10.1002/int.22593
    [18] Ibrahim A, Ismail A, Juahir H, et al. Water quality modelling using principal component analysis and artificial neural network[J]. Marine Pollution Bulletin, 2023, 187: 114493. doi: 10.1016/j.marpolbul.2022.114493
    [19] Mai Gengchen, Xuan Yao, Zuo Wenyun, et al. Sphere2Vec: a general-purpose location representation learning over a spherical surface for large-scale geospatial predictions[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2023, 202: 439−462. doi: 10.1016/j.isprsjprs.2023.06.016
    [20] Scarselli F, Gori M, Tsoi A C, et al. The graph neural network model[J]. IEEE Transactions on Neural Networks, 2009, 20(1): 61−80. doi: 10.1109/TNN.2008.2005605
    [21] Wang Xiaoqi, Yen K, Hu Yifan, et al. DeepGD: a deep learning framework for graph drawing using GNN[J]. IEEE Computer Graphics and Applications, 2021, 41(5): 32−44. doi: 10.1109/MCG.2021.3093908
    [22] Wang Dujuan, Zhu Jiacheng, Yin Yunqiang, et al. Dynamic travel time prediction with spatiotemporal features: using a GNN-based deep learning method[J]. Annals of Operations Research, 2024, 340(1): 571−591. doi: 10.1007/s10479-023-05260-2
    [23] Kipf T N, Welling M. Semi-supervised classification with graph convolutional networks[C]//5th International Conference on Learning Representations. Toulon, 2017. (查阅网上资料, 未找到本条文献母体文献及出版信息, 请确认)
    [24] Derr T, Ma Yao, Tang Jiliang. Signed graph convolutional networks[C]//2018 IEEE International Conference on Data Mining (ICDM). Singapore: IEEE, 2018: 929−934.
    [25] Polamuri S R, Srinivas D K, Mohan D A K. Multi-model generative adversarial network hybrid prediction algorithm (MMGAN-HPA) for stock market prices prediction[J]. Journal of King Saud University-Computer and Information Sciences, 2022, 34(9): 7433−7444. doi: 10.1016/j.jksuci.2021.07.001
    [26] Zhang Jianguang, Zhang Xuyang, Yang Jianfeng, et al. Deep LSTM and GAN based short-term load forecasting method at the zone level[C]//2020 International Conference on Artificial Intelligence in Information and Communication (ICAIIC). Fukuoka: IEEE, 2020: 613−618.
    [27] Festag S, Spreckelsen C. Medical multivariate time series imputation and forecasting based on a recurrent conditional Wasserstein GAN and attention[J]. Journal of Biomedical Informatics, 2023, 139: 104320. doi: 10.1016/j.jbi.2023.104320
    [28] Gulrajani I, Ahmed F, Arjovsky M, et al. Improved training of wasserstein GANS[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach: Curran Associates Inc. , 2017: 5769−5779.
    [29] Zhang Chenhan, Yu J J Q, Liu Yi. Spatial-temporal graph attention networks: a deep learning approach for traffic forecasting[J]. IEEE Access, 2019, 7: 166246−166256. doi: 10.1109/ACCESS.2019.2953888
    [30] Fang Fang, Zhang Pengpeng, Zhou Bo, et al. Atten-GAN: pedestrian trajectory prediction with GAN based on attention mechanism[J]. Cognitive Computation, 2022, 14(6): 2296−2305. doi: 10.1007/s12559-022-10029-z
    [31] Bynagari N B. GANs trained by a two time-scale update rule converge to a local Nash equilibrium[J]. Asian Journal of Applied Science and Engineering, 2019, 8(1): 25−34. doi: 10.18034/ajase.v8i1.9
    [32] Zhang Kang, Zhong Guoqiang, Dong Junyu, et al. Stock market prediction based on generative adversarial network[J]. Procedia Computer Science, 2019, 147: 400−406. doi: 10.1016/j.procs.2019.01.256
    [33] Zha Wenshu, Liu Yuping, Wan Yujin, et al. Forecasting monthly gas field production based on the CNN-LSTM model[J]. Energy, 2022, 260: 124889. doi: 10.1016/j.energy.2022.124889
    [34] Yu Jingxin, Zhang Xin, Xu Linlin, et al. A hybrid CNN-GRU model for predicting soil moisture in maize root zone[J]. Agricultural Water Management, 2021, 245: 106649. doi: 10.1016/j.agwat.2020.106649
  • 加载中
图(9) / 表(4)
计量
  • 文章访问数:  31
  • HTML全文浏览量:  13
  • PDF下载量:  9
  • 被引次数: 0
出版历程
  • 收稿日期:  2024-09-28
  • 修回日期:  2024-12-18
  • 网络出版日期:  2025-01-08

目录

    /

    返回文章
    返回