机器学习:海底矿产资源智能勘探的新途径

doi:10.13745/j.esf.sf.2023.5.90

地学前缘 ›› 2024, Vol. 31 ›› Issue (3): 520-529.DOI: 10.13745/j.esf.sf.2023.5.90

机器学习:海底矿产资源智能勘探的新途径

刘洋¹^,²(), 李三忠¹^,², 钟世华¹^,²^,^*(), 郭广慧¹^,², 刘嘉情¹^,², 牛警徽¹^,², 薛梓萌¹^,², 周建平¹^,², 董昊¹^,², 索艳慧¹^,²

1.中国海洋大学海洋地球科学学院深海圈层与地球系统教育部前沿科学中心/海底科学与探测技术教育部重点实验室, 山东青岛 266100
2.青岛海洋科学与技术国家实验室海洋矿产资源评价与探测技术功能实验室, 山东青岛 266237

收稿日期:2022-11-23 修回日期:2023-04-13 出版日期:2024-05-25 发布日期:2024-05-25
通信作者: *钟世华(1989—),男,副教授,主要从事矿床学和地质大数据研究。E-mail: zhongshihua@ouc.edu.cn
作者简介:刘洋(1997—),男,博士研究生,海洋地质专业。E-mail: liuyang9553@stu.ouc.edu.cn
基金资助:
青岛海洋科学与技术试点国家实验室山东省专项(2022QNLM05032);中央高校基本科研业务费专项(202172002);国家自然科学基金项目(42203066);山东省自然科学基金项目(ZR2020QD027)

Machine learning: A new approach to intelligent exploration of seafloor mineral resources

LIU Yang¹^,²(), LI Sanzhong¹^,², ZHONG Shihua¹^,²^,^*(), GUO Guanghui¹^,², LIU Jiaqing¹^,², NIU Jinghui¹^,², XUE Zimeng¹^,², ZHOU Jianping¹^,², DONG Hao¹^,², SUO Yanhui¹^,²

1. MOE Frontiers Science Center for Deep Ocean Multispheres and Earth System/MOE Key Lab of Submarine Geosciences and Prospecting Techniques, College of Marine Geosciences, Ocean University of China, Qingdao 266100, China
2. Laboratory for Marine Mineral Resources, National Laboratory for Marine Science and Technology (Qingdao), Qingdao 266237, China

Received:2022-11-23 Revised:2023-04-13 Online:2024-05-25 Published:2024-05-25

摘要/Abstract

摘要：

海底蕴藏着丰富的关键矿产资源,是当前研究的热点,也是未来产业新领域。随着海洋探测技术的不断进步,海底矿产勘探的数据量和数据维数急剧增加,给数据处理与解释带来了巨大困难和挑战。面对海量数据,传统的数据解释与分析方法暴露出许多问题。机器学习以其强大的自学能力,为无法解决或难以解决的问题提供了一系列智能分析决策方案,提高了数据分析的效率,是海底矿产资源智能勘探的新途径。近年来,机器学习在地球科学领域获得了广泛的关注和研究。为此,围绕机器学习技术应用于海底资源勘探技术,本文首先简要介绍了机器学习中经典的模型算法,然后详细阐述了机器学习在海底能源矿产和金属矿产两个方面的应用现状,最后总结了机器学习在海底矿产智能勘探领域的应用前景,指出了现有研究中存在的问题,提出了解决方案和未来的发展方向。

关键词: 机器学习, 铁锰结核, 富钴结壳, 天然气水合物, 海底矿产

Abstract:

The seafloor is characterized by abundant key mineral resources, which is a hotspot of current research and a new field of industry in the future. With the continuous progress of ocean exploration technology, the volume and dimensions of data from seafloor mineral exploration have increased dramatically, which has brought great difficulties and challenges to data processing and interpretation. In the face of massive data, traditional data interpretation and analysis methods expose many problems. Machine learning, with its strong self-learning ability, provides a series of intelligent analysis and decision-making solutions for unsolvable or difficult problems, improving the efficiency of data analysis. It is a new way for the intelligent exploration of subsea mineral resources. In recent years, machine learning has obtained extensive attention and research in the field of geosciences. Therefore, focusing on the application of machine learning to seafloor resource exploration technology, this paper firstly briefly introduces the classical model algorithms in machine learning; then elaborates on the application status of machine learning in two aspects of seafloor energy resources and metal mineral, and finally summarizes the application prospects of machine learning in the field of intelligent exploration of seafloor minerals, points out the problems in existing research, and proposes solutions and future development directions.

Key words: machine learning, ferromanganese nodule, cobalt-rich crusts, natural gas hydrate, seafloor minerals

中图分类号:

刘洋, 李三忠, 钟世华, 郭广慧, 刘嘉情, 牛警徽, 薛梓萌, 周建平, 董昊, 索艳慧. 机器学习:海底矿产资源智能勘探的新途径[J]. 地学前缘, 2024, 31(3): 520-529.

LIU Yang, LI Sanzhong, ZHONG Shihua, GUO Guanghui, LIU Jiaqing, NIU Jinghui, XUE Zimeng, ZHOU Jianping, DONG Hao, SUO Yanhui. Machine learning: A new approach to intelligent exploration of seafloor mineral resources[J]. Earth Science Frontiers, 2024, 31(3): 520-529.

图/表 5

图1 机器学习方法及其应用(引自文献[10])

Fig.1 Machine learning methods and their applications. Adapted from [10].

图2 有监督机器学习方法的示意 (a)逻辑回归;(b)支持向量机;(c)随机森林;(d)深度神经网络。

Fig.2 Schematic representation of supervised machine learning methods (a) Logistic regression; (b) Support vector machine; (c) Random forest; (d) Deep neural network.

图3 无监督机器学习方法的示意 (a)K均值聚类;(b)自编码器。

Fig.3 Schematic representation of unsupervised machine learning methods (a) K-means clustering; (b) Auto-encoder.

图4 机器学习在海底非金属矿产中的应用(引自文献[31⇓⇓-34])

Fig.4 Application of machine learning to seabed non-metallic minerals. Adapted from [31⇓⇓-34].

图5 机器学习在海底金属矿产中的应用(引自文献[43⇓⇓⇓-47])

Fig.5 Application of machine learning in seabed metal minerals. Adapted from [43⇓⇓⇓-47].

参考文献 58

[1]	PETERSEN S. Marine mineral resources[M]//HARFF J, MESCHEDE M, PETERSEN S, et al. Encyclopedia of marine geosciences. Dordrecht: Springer, 2016: 475-480.
[2]	PETERSEN S, KRÄTSCHELL A, AUGUSTIN N, et al. News from the seabed-geological characteristics and resource potential of deep-sea mineral resources[J]. Marine Policy, 2016, 70: 175-187.
[3]	RONA P A. The changing vision of marine minerals[J]. Ore Geology Reviews, 2007, 33(3): 618-666.
[4]	石学法, 符亚洲, 李兵, 等. 我国深海矿产研究: 进展与发现(2011—2020)[J]. 矿物岩石地球化学通报, 2021, 40(2): 305-318.
[5]	周守为, 李清平, 朱海山, 等. 海洋能源勘探开发技术现状与展望[J]. 中国工程科学, 2016, 18(2): 19-31.
[6]	HEIN J R, KOSCHINSKY A, KUHN T. Deep-ocean polymetallic nodules as a resource for critical materials[J]. Nature Reviews Earth & Environment, 2020, 1(3): 158-169.
[7]	杨燕子, 陈华勇. 大洋富钴结壳研究进展及展望[J]. 大地构造与成矿学, 2023, 47(1): 1-19.
[8]	HANNINGTON M, JAMIESON J, MONECKE T, et al. The abundance of seafloor massive sulfide deposits[J]. Geology, 2011, 39(12): 1155-1158.
[9]	KVENVOLDEN K. Gas hydrates: geological perspective and global change[J]. Reviews of Geophysics, 1993, 31(2): 173-187.
[10]	BERGEN K J, JOHNSON P A, DE HOOP M V, et al. Machine learning for data-driven discovery in solid Earth geoscience[J]. Science, 2019, 363(6433): eaau0323.
[11]	DAI H C, MACBETH C. Automatic picking of seismic arrivals in local earthquake data using an artificial neural network[J]. Geophysical Journal International, 1995, 120(3): 758-774.
[12]	DOWLA F, TAYLOR S R, ANDERSON R W. Seismic discrimination with artificial neural networks: preliminary results with regional spectral data[J]. Bulletin of the Seismological Society of America, 1990, 80(5): 1346-1373.
[13]	DYSART P, PULLI J. Regional seismic event classification at the NORESS array: seismological measurements and the use of trained neural networks[J]. Bulletin of the Seismological Society of America, 1990, 80(6B): 1910-1933.
[14]	周永章, 左仁广, 刘刚, 等. 数学地球科学跨越发展的十年: 大数据、人工智能算法正在改变地质学[J]. 矿物岩石地球化学通报, 2021, 40(3): 556-573, 777.
[15]	周永章, 张良均, 张奥多, 等. 地球科学大数据挖掘与机器学习[M]. 广州: 中山大学出版社, 2018.
[16]	REICHSTEIN M, CAMPS-VALLS G, STEVENS B, et al. Deep learning and process understanding for data-driven Earth system science[J]. Nature, 2019, 566(7743): 195-204.
[17]	COX D R. The regression analysis of binary sequences[J]. Journal of the Royal Statistical Society: Series B (Methodological), 1959, 21(1): 238.
[18]	CORTES C, VAPNIK V. Support-vector networks[J]. Machine Learning, 1995, 20(3): 273-297.
[19]	VAPNIK V N. An overview of statistical learning theory[J]. IEEE Transactions on Neural Networks, 1999, 10(5): 988-999. DOI PMID
[20]	BURGES C J C. A tutorial on support vector machines for pattern recognition[J]. Data Mining and Knowledge Discovery, 1998, 2(2): 121-167.
[21]	CHANG C C, LIN C J. LIBSVM[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3): 1-27.
[22]	BREIMAN L. Random forests[J]. Machine Learning, 2001, 45(1): 5-32.
[23]	BREIMAN L. Bagging predictors[J]. Machine Learning, 1996, 24(2): 123-140.
[24]	LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444.
[25]	LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.
[26]	ABDEYAZDAN M. Data clustering based on hybrid K-harmonic means and modifier imperialist competitive algorithm[J]. The Journal of Supercomputing, 2014, 68(2): 574-598.
[27]	FRADKIN D. Within-class and unsupervised clustering improve accuracy and extract local structure for supervised classification[D]. New Brunswick: Rutgers, The State University of New Jersey, 2006.
[28]	GOODFELLOW I, BENGIO Y, COURVILLE A. Deep learning[M]. Cambridge: The MIT Press, 2016.
[29]	BOURLARD H, KAMP Y. Auto-association by multilayer perceptrons and singular value decomposition[J]. Biological Cybernetics, 1988, 59(4): 291-294.
[30]	ZHAI J H, ZHANG S F, CHEN J F, et al. Autoencoder and its various variants[C]//Proceedings of 2018 IEEE international conference on systems, man, and cybernetics (SMC). Miyazaki: IEEE, 2018: 415-419.
[31]	LIM B, YU H, YOON D, et al. Machine learning derived AVO analysis on marine 3D seismic data over gas reservoirs near South Korea[J]. Journal of Petroleum Science and Engineering, 2021, 197: 108105.
[32]	ZHU L Q, WEI J G, WU S G, et al. Application of unlabelled big data and deep semi-supervised learning to significantly improve the logging interpretation accuracy for deep-sea gas hydrate-bearing sediment reservoirs[J]. Energy Reports, 2022, 8: 2947-2963.
[33]	MONDAL I, SINGH H K. Core-log integration and application of machine learning technique for better reservoir characterisation of Eocene carbonates, Indian offshore[J]. Energy Geoscience, 2022, 3(1): 49-62.
[34]	YOU J C, CAO J X, WANG X J, et al. Shear wave velocity prediction based on LSTM and its application for morphology identification and saturation inversion of gas hydrate[J]. Journal of Petroleum Science and Engineering, 2021, 205: 109027.
[35]	PANG Y M, SHI B B, GUO X W, et al. Source-reservoir relationships and hydrocarbon charging history in the central uplift of the South Yellow Sea Basin (East Asia): constrained by machine learning procedure and basin modeling[J]. Marine and Petroleum Geology, 2021, 123: 104731.
[36]	HAN F L, ZHANG H B, RUI J W, et al. Multiple point geostatistical simulation with adaptive filter derived from neural network for sedimentary facies classification[J]. Marine and Petroleum Geology, 2020, 118: 104406.
[37]	BROWN N, ROUBÍČKOVÁ A, LAMPAKI I, et al. Machine learning on Crays to optimize petrophysical workflows in oil and gas exploration[J]. Concurrency and Computation: Practice and Experience, 2020, 32(20): e5655.
[38]	MUKHERJEE B, SAIN K. Prediction of reservoir parameters in gas hydrate sediments using artificial intelligence (AI): a case study in Krishna-Godavari Basin (NGHP Exp-02)[J]. Journal of Earth System Science, 2019, 128(7): 199.
[39]	CHONG L, SINGH H, CREASON C G, et al. Application of machine learning to characterize gas hydrate reservoirs in Mackenzie Delta (Canada) and on the Alaska north slope (USA)[J]. Computational Geosciences, 2022, 26(5): 1151-1165.
[40]	SINGH H, SEOL Y, MYSHAKIN E M. Prediction of gas hydrate saturation using machine learning and optimal set of well-logs[J]. Computational Geosciences, 2021, 25(1): 267-283.
[41]	LI C, LIU X. Research on the estimate of gas hydrate saturation based on LSTM recurrent neural network[J]. Energies, 2020, 13(24): 6536.
[42]	LEE J, BYUN J, KIM B, et al. Delineation of gas hydrate reservoirs in the Ulleung Basin using unsupervised multi-attribute clustering without well log data[J]. Journal of Natural Gas Science and Engineering, 2017, 46: 326-337.
[43]	DUTKIEWICZ A, JUDGE A, MÜLLER R D. Environmental predictors of deep-sea polymetallic nodule occurrence in the global ocean[J]. Geology, 2020, 48(3): 293-297.
[44]	HARI V N, KALYAN B, CHITRE M, et al. Spatial modeling of deep-sea ferromanganese nodules with limited data using neural networks[J]. IEEE Journal of Oceanic Engineering, 2018, 43(4): 997-1014.
[45]	JULIANI C, JULIANI E. Deep learning of terrain morphology and pattern discovery via network-based representational similarity analysis for deep-sea mineral exploration[J]. Ore Geology Reviews, 2021, 129: 103936.
[46]	MIMURA K, NAKAMURA K, TAKAO K, et al. Automated detection of hydrothermal emission signatures from multi-beam echo sounder images using deep learning[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16: 2703-2710.
[47]	GAZIS I Z, GREINERT J. Importance of spatial autocorrelation in machine learning modeling of polymetallic nodules, model uncertainty and transferability at local scale[J]. Minerals, 2021, 11(11): 1172.
[48]	SCHOENING T, KUHN T, NATTKEMPER T. Estimation of poly-metallic nodule coverage in benthic images[C]//Proceedings of the 41st conference of the Underwater Mining Institute (UMI). Shanghai: UMI, 2012: 15-20.
[49]	邓君兰, 董澧辉, 宋伟, 等. 基于Sea-thru和Mask R-CNN的深海多金属结核图像处理[J]. 矿冶工程, 2022, 42(2): 9-13.
[50]	PRABHAKARAN K, VARSHNEY N, RAMESH R, et al. Underwater image and video processing to detect polymetallic nodule abundance using haar-cascade and template matching feature[C]//Proceedings of OCEANS 2022-Chennai. Chennai: IEEE, 2022: 1-6.
[51]	GAZIS I Z, SCHOENING T, ALEVIZOS E, et al. Quantitative mapping and predictive modeling of Mn nodules’ distribution from hydroacoustic and optical AUV data linked by random forests machine learning[J]. Biogeosciences, 2018, 15(23): 7347-7377.
[52]	HU G, ZHAO H M, HAN F L, et al. Recognition of cobalt-rich crusts based on multi-classifier fusion in seafloor mining environments[J]. Marine Georesources & Geotechnology, 2021, 39(10): 1205-1214.
[53]	HONG F, HUANG M Y, FENG H H, et al. First demonstration of recognition of manganese crust by deep-learning networks with a parametric acoustic probe[J]. Minerals, 2022, 12(2): 249.
[54]	NEETTIYATH U, SATO T, SANGEKAR M, et al. Identification of manganese crusts in 3D visual reconstructions to filter geo-registered acoustic sub-surface measurements[C]//Proceedings of OCEANS 2015-MTS/IEEE Washington. Washington: IEEE, 2015: 1-6.
[55]	NEETTIYATH U, THORNTON B, SANGEKAR M, et al. An AUV based method for estimating hectare-scale distributions of deep sea cobalt-rich manganese crust deposits[C]//Proceedings of OCEANS 2019-Marseille. Marseille: IEEE, 2019: 1-6.
[56]	LIU L S, LU J L, TAO C H, et al. Fuzzy forest machine learning predictive model for mineral prospectivity: a case study on southwest Indian ridge 48. 7°E-50. 5°E[J]. Natural Resources Research, 2022, 31(1): 99-116.
[57]	ZHANG X N, SONG S J, LI J B, et al. Robust LS-SVM regression for ore grade estimation in a seafloor hydrothermal sulphide deposit[J]. Acta Oceanologica Sinica, 2013, 32(8): 16-25.
[58]	ZHANG X N, SONG S J, WU C. Robust Bayesian classification with incomplete data[J]. Cognitive Computation, 2013, 5(2): 170-187.

机器学习:海底矿产资源智能勘探的新途径

Machine learning: A new approach to intelligent exploration of seafloor mineral resources

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 5

参考文献 58

相关文章 15

编辑推荐

Metrics

本文评价

[1]	张焕宝, 贺海洋, 杨仕教, 李亚林, 毕文军, 韩世礼, 郭钦鹏, 杜青. 基于机器学习的埃达克质岩构造背景判别研究[J]. 地学前缘, 2024, 31(4): 417-428.
[2]	张利军, 鲁文豪, 张建东, 彭光雄, 卜建财, 唐凯, 谢渐成, 徐质彬, 杨海燕. 基于深度学习的镜下岩石、矿物薄片识别[J]. 地学前缘, 2024, 31(3): 498-510.
[3]	苏恺明, 徐耀辉, 徐旺林, 张月巧, 白斌, 李阳, 严刚. 鄂尔多斯盆地延长组多油源贡献比例与分布规律:基于机器学习与可解释性研究[J]. 地学前缘, 2024, 31(3): 530-540.
[4]	王子烨, 左仁广. 基于多源数据融合的喜马拉雅淡色花岗岩识别[J]. 地学前缘, 2023, 30(5): 216-226.
[5]	宋轩宇, 许民, 康世昌, 孙立平. 基于机器学习的冰冻圈典型流域水文过程模拟研究[J]. 地学前缘, 2023, 30(4): 451-469.
[6]	朱紫怡, 周飞, 王瑀, 周统, 侯照亮, 邱昆峰. 基于机器学习的锆石成因分类研究[J]. 地学前缘, 2022, 29(5): 464-475.
[7]	孙涛, 吴涛, 葛阳, 樊奇, 李丽霞, 吕鑫. 琼东南盆地深水区浅表层水合物稀有气体地球化学特征及意义[J]. 地学前缘, 2022, 29(5): 476-482.
[8]	贾永刚, 阮文凤, 胡乃利, 乔玥, 李正辉, 胡聪. 现代暖期气候变暖对南海北部陆坡天然气水合物分解潜在影响[J]. 地学前缘, 2022, 29(4): 191-201.
[9]	胡义明, 陈腾, 罗序义, 唐超, 梁忠民. 基于机器学习模型的淮河流域中长期径流预报研究[J]. 地学前缘, 2022, 29(3): 284-291.
[10]	姚会强, 刘永刚, 张伙带, 梁东红, 任江波, 于淼, 邓希光, 何高文. 维嘉平顶海山富钴铁锰结壳空间分布特征:基于“蛟龙号”载人潜水器近海底观测资料的分析[J]. 地学前缘, 2021, 28(6): 331-342.
[11]	左仁广. 基于数据科学的矿产资源定量预测的理论与方法探索[J]. 地学前缘, 2021, 28(3): 49-55.
[12]	张振杰, 成秋明, 杨玠, 武国朋, 葛云钊. 机器学习与成矿预测:以闽西南铁多金属矿预测为例[J]. 地学前缘, 2021, 28(3): 221-235.
[13]	任江波, 邓义楠, 赖佩欣, 何高文, 王汾连, 姚会强, 邓希光, 刘永刚. 太平洋调查区多金属结核的地球化学特征和成因[J]. 地学前缘, 2021, 28(2): 412-425.
[14]	罗建民，张旗. 大数据开创地学研究新途径：查明相关关系，增强研究可行性[J]. 地学前缘, 2019, 26(4): 6-12.
[15]	章宝月，孙建鹍，罗熊，金维浚，王龙，杜雪亮，陈万峰，杜君，张旗，朱月琴. 三类构造背景辉长岩单斜辉石主量元素和微量元素的数据分析研究[J]. 地学前缘, 2019, 26(4): 33-44.