遗传算法在瞬变电磁深度学习反演中的优化策略

吴文宇; 张莹莹; 吴新宇; 谢斌

doi:10.13568/j.cnki.651094.651316.2025.03.31.0005

您当前的位置：

首页 >

文章列表页 >

遗传算法在瞬变电磁深度学习反演中的优化策略

新能源与绿色生产·庆祝新疆维吾尔自治区成立70周年专题 | 更新时间：2026-01-23

- 遗传算法在瞬变电磁深度学习反演中的优化策略
- 遗传算法在瞬变电磁深度学习反演中的优化策略
- 新疆大学学报（自然科学版中英文） 2025年42卷第3期页码：280-299
- 作者机构：
  
  新疆大学地质与矿业工程学院
- 作者简介：
- 基金信息：
  
  新疆维吾尔自治区重点研发计划“新疆天山及邻区早二叠世金镍重大成矿事件与找矿评价关键技术开发”（2024B03005-3）
- DOI：10.13568/j.cnki.651094.651316.2025.03.31.0005
  中图分类号： P631.325
- 纸质出版：2025
- 稿件说明：
移动端阅览
[1]吴文宇,张莹莹,吴新宇,等.遗传算法在瞬变电磁深度学习反演中的优化策略[J].新疆大学学报(自然科学版中英文),2025,42(03):280-299.
[1]吴文宇,张莹莹,吴新宇,等.遗传算法在瞬变电磁深度学习反演中的优化策略[J].新疆大学学报(自然科学版中英文),2025,42(03):280-299. DOI： 10.13568/j.cnki.651094.651316.2025.03.31.0005.

DOI：10.13568/j.cnki.651094.651316.2025.03.31.0005.

摘要

瞬变电磁一维反演技术已在地质工程中得到了广泛应用，但是该方法较为依赖初始模型，抗噪能力较差，难以实现实时反演．针对上述问题，采用卷积长短期记忆(CNN-LSTM)混合神经网络，利用回线源瞬变电磁一维正演程序得到的采样时间-衰减电压作为网络输入数据，同时采用Adam优化器与ReduceLROnPlateau学习率调度器相结合的优化策略自适应调整学习率．针对网络结构超参数设置依赖经验值、缺乏科学性导致的算力以及时间浪费问题，在模型训练阶段采用遗传算法(GA)对神经网络结构进行超参数寻优，提高模型性能．最终，在输出层得到与输入的电磁响应数据对应的深度-电阻率数据，实现瞬变电磁数据深度学习反演．利用训练后的GA-CNN-LSTM网络对随机生成的3层模型以及5层模型进行实时预测，测试集评价指标R2>0.9，验证了所提算法的可靠性．进一步对加噪数据进行反演，完成训练的神经网络对4种常见模型的平均反演用时为0.13 s，平均反演结构相似度达90.138%，两项指标均优于Occam以及LSTM反演方法．进一步对三维正演模型进行数据反演，验证了所提算法的泛化能力．最后，对实测数据分别进行了Occam反演和神经网络反演，在保证反演精度的同时，完成训练的神经网络仅用时0.73 s，表明所提算法具有实用性．

Abstract

Transient electromagnetic(TEM) 1D inversion has been widely applied in geological engineering

yet these conventional methods remain constrained by strong dependence on initial models

limited noise resistance

and inefficiency in achieving real-time inversion. To address these challenges

we propose a convolutional neural network-long short-term memory hybrid architecture tailored to the inherent characteristics of TEM inversion.Using loop-source TEM 1D forward modeling

we generate training data comprising sampling time-decay voltage pairs as network inputs. An optimization strategy combining the Adam optimizer with the ReduceLROnPlateau learning rate scheduler is implemented to adaptively adjust learning rates during parameter updates. In view of the problem that the hyper-parameter setting of the current network structure depends on the empirical value and lacks scientificity

which leads to the waste of computing power and time

the genetic algorithm is proposed to optimize the hyper-parameters of the neural network structure in the model training stage to reduce the training cost and improve the model performance. The output layer provides subsurface resistivity-depth profiles corresponding to input electromagnetic responses

enabling deep learning-based TEM inversion. The trained GA-CNN-LSTM network demonstrates robust performance in real-time predictions for randomly generated three-layer and fivelayer models

with validation metrics yielding R2>0.9. Further evaluation using noise-contaminated data reveals that the optimized network achieves an average inversion time of 0.13 s and a structural similarity index of 90.138%across four common models

outperforming both Occam and LSTM inversion methods. Generalization capability is validated through successful inversion of 3D forward modeling data. These results demonstrate the algorithm's reliability

computational efficiency

and practical utility in complex geological scenarios. Finally

Occam inversion and neural network inversion are carried out on the measured data respectively. The trained neural network only take 0.73 s to complete the inversion accuracy

which verifies the practicability of the algorithm in this paper.

关键词

Keywords

references

滕吉文．强化开展地壳内部第二深度空间金属矿产资源地球物理找矿、勘探和开发[J]．地质通报，2006,25(7):767-771.TENG J W.Strengthening geophysical exploration and exploitation of metallic minerals in the second deep space of the crustal interior[J].Geological Bulletin of China,2006,25(7):767-771.(in Chinese)

董树文，李廷栋，高锐，等．地球深部探测国际发展与我国现状综述[J]．地质学报，2010,84(6):743-770.DONG S W,LI T D,GAO R,et al.International progress in probing the earth’s lithosphere and deep interior:A review[J].Acta Geologica Sinica,2010,84(6):743-770.(in Chinese)

刘光鼎．推动地球物理方法创新，引领探测仪器技术未来[J]．地球物理学报，2017,60(11):4145-4148.LIU G D.Promote the innovation of geophysical methodology,and lead the future of exploration apparatus technology[J].Chinese Journal of Geophysics,2017,60(11):4145-4148.(in Chinese)

底青云，朱日祥，薛国强，等．我国深地资源电磁探测新技术研究进展[J]．地球物理学报，2019,62(6):2128-2138.DI Q Y,ZHU R X,XUE G Q,et al.New development of the electromagnetic(EM)methods for deep exploration[J].Chinese Journal of Geophysics,2019,62(6):2128-2138.(in Chinese)

滕吉文．高精度地球物理学是创新未来的必然发展轨迹[J]．地球物理学报，2021,64(4):1131-1144.TENG J W.High-precision geophysics:The inevitable development track of the innovative future[J].Chinese Journal of Geophysics,2021,64(4):1131-1144.(in Chinese)

NABIGHIAN M N.Electromagnetic methods in applied geophysics[M]//NEITZEL E B.Investigations in geophysics.Tulsa:Society of Exploration Geophysicists,1987.

李貅．瞬变电磁测深的理论与应用[M]．西安：陕西科学技术出版社，2002.LI X.Theory and application of transient electromagnetic sounding[M].Xi’an:Shaanxi Science and Technology Press,2002.(in Chinese)

STRACK K M.Exploration with deep transient electromagnetics[M]//SHARMA P V.Methods in geochemistry and geophysics.Amsterdam:Elsevier,1992.

OLDENBURG D W,LI Y G,FARQUHARSON C G,et al.Applications of geophysical inversions in mineral exploration[J].The Leading Edge,1998,17(4):461-465.

GUO Z W,XUE G Q,LIU J X,et al.Electromagnetic methods for mineral exploration in China:A review[J].Ore Geology Reviews,2020,118:103357.

CHANG J H,SU B Y,MALEKIAN R,et al.Detection of water-filled mining goaf using mining transient electromagnetic method[J].IEEE Transactions on Industrial Informatics,2020,16(5):2977-2984.

薛国强，武欣，陈卫营，等．接地源短偏移瞬变电磁法理论及其在深部矿产勘探中的应用[J]．中国科学：地球科学，2025,55(2):642-654.XUE G Q,WU X,CHEN W Y,et al.Grounded-source short offset transient electromagnetic method:Theory and applications in deep mineral exploration[J].Science China:Earth Sciences,2025,55(2):642-654.(in Chinese)

DANIELSEN J E,AUKEN E,JORGENSEN F,et al.The application of the transient electromagnetic method in hydrogeophysical surveys[J].Journal of Applied Geophysics,2003,53(4):181-198.

于景邨，刘志新，刘树才，等．深部采场突水构造矿井瞬变电磁法探查理论及应用[J]．煤炭学报，2007,32(8):818-821.YU J C,LIU Z X,LIU S C,et al.Theoretical analysis of mine transient electromagnetic method and its application in detecting water burst structures in deep coal stope[J].Journal of China Coal Society,2007,32(8):818-821.(in Chinese)

姚伟华，王鹏，李明星，等．三分量地-孔瞬变电磁法积水采空区探测试验[J]．煤田地质与勘探，2019,47(5):54-62.YAO W H,WANG P,LI M X,et al.Experimental study of three-component down-hole TEM for detecting water-filled goaf[J].Coal Geology&Exploration,2019,47(5):54-62.(in Chinese)

AUKEN E,VIOLETTE S,D’OZOUVILLE N,et al.An integrated study of the hydrogeology of volcanic islands using helicopter borne transient electromagnetic:Application in the Galapagos Archipelago[J].Comptes Rendus Geoscience,2009,341(10/11):899-907.

FINN C A,BEDROSIAN P A,HOLBROOK W S,et al.Geophysical imaging of the Yellowstone hydrothermal plumbing system[J].Nature,2022,603(7902):643-647.

LIN J,CHEN J,ZHANG Y.Rapid and high-resolution detection of urban underground space using transient electromagnetic method[J].IEEE Transactions on Industrial Informatics,2022,18(4):2622-2631.

方文藻，李予国，李貅．瞬变电磁测深法原理[M]．西安：西北工业大学出版社，1993.FANG W Z,LI Y G,LI X.Principle of transient electromagnetic sounding method[M].Xi’an:Northwestern Polytechnical University Press,1993.(in Chinese)

KELLER G V,PRITCHARD J I,JACOBSON J J,et al.Megasource time-domain electromagnetic sounding methods[J].Geophysics,1984,49(7):993-1009.

薛国强，李貅，底青云．瞬变电磁法理论与应用研究进展[J]．地球物理学进展，2007,22(4):1195-1200.XUE G Q,LI X,DI Q Y.The progress of TEM in theory and application[J].Progress in Geophysics,2007,22(4):1195-1200.(in Chinese)

OLDENBURG D W,HABER E,SHEKHTMAN R.Three dimensional inversion of multisource time domain electromagnetic data[J].Geophysics,2013,78(1):E47-E57.

万玲，林婷婷，林君，等．基于自适应遗传算法的MRS-TEM联合反演方法研究[J]．地球物理学报，2013,56(11):3728-3740.WAN L,LIN T T,LIN J,et al.Joint inversion of MRS and TEM data based on adaptive genetic algorithm[J].Chinese Journal of Geophysics,2013,56(11):3728-3740.(in Chinese)

李明星，肖林通，张倚瑞，等．瞬变电磁粒子群优化反演研究[J]．煤炭技术，2014,33(9):302-304.LI M X,XIAO L T,ZHANG Y R,et al.Research on particle swarm optimization inversion of transient electromagnetic method[J].Coal Technology,2014,33(9):302-304.(in Chinese)

程久龙，李明星，肖艳丽，等．全空间条件下矿井瞬变电磁法粒子群优化反演研究[J]．地球物理学报，2014,57(10):3478-3484.CHENG J L,LI M X,XIAO Y L,et al.Study on particle swarm optimization inversion of mine transient electromagnetic method in whole-space[J].Chinese Journal of Geophysics,2014,57(10):3478-3484.(in Chinese)

LI R Y,ZHANG H Q,YU N,et al.A fast approximation for 1-D inversion of transient electromagnetic data by using a back propagation neural network and improved particle swarm optimization[J].Nonlinear Processes in Geophysics,2019,26(4):445-456.

孙怀凤，张诺亚，柳尚斌，等．基于L1范数的瞬变电磁非线性反演[J]．地球物理学报，2019,62(12):4860-4873.SUN H F,ZHANG N Y,LIU S B,et al.L1-norm based nonlinear inversion of transient electromagnetic data[J].Chinese Journal of Geophysics,2019,62(12):4860-4873.(in Chinese)

XUE G Q,LI H,HE Y M,et al.Development of the inversion method for transient electromagnetic data[J].IEEE Access,2020,8:146172-146181.

胡琪鑫，徐亚．地球物理信号特征识别与解释的机器学习方法及应用综述[J]．地球物理学进展，2022,37(6):2395-2407.HU Q X,XU Y.Review of machine learning and application of geophysical signal feature recognition and interpretation[J].Progress in Geophysics,2022,37(6):2395-2407.(in Chinese)

李创社，张彦鹏，李实，等．瞬变电磁勘探中的人工神经网络反演法[J]．西安交通大学学报，2001,35(6):604-607+615.LI C S,ZHANG Y P,LI S,et al.New inversion method of artificial neural network in transient electromagnetic inversion[J].Journal of Xi’an Jiaotong University,2001,35(6):604-607+615.(in Chinese)

王秀臣．基于人工神经网络的瞬变电磁反演解释及应用研究[D]．西安：西北大学，2006.WANG X C.Study on inversion and interpretation of the transient electromagnetic method based on artificial neural network and its applications[D].Xi’an:Northwest University,2006.(in Chinese)

秦善强，付志红，朱学贵，等．遗传神经网络的瞬变电磁视电阻率求解算法[J]．电工技术学报，2017,32(12):146-154+250.QIN S Q,FU Z H,ZHU X G,et al.Genetic neural network for apparent resistivity solution of transient electromagnetic[J].Transactions of China Electrotechnical Society,2017,32(12):146-154+250.(in Chinese)

COLOMBO D,TURKOGLU E,LI W C,et al.Physics-driven deep-learning inversion with application to transient electromagnetics[J].Geophysics,2021,86(3):E209-E224.

PUZYREV V,SWIDINSKY A.Inversion of 1D frequency-and time-domain electromagnetic data with convolutional neural networks[J].Computers&Geosciences,2021,149:104681.

冼锦炽，蔡红柱，熊咏春，等．基于深度学习的地面拖曳式瞬变电磁快速成像方法[J]．工程地球物理学报，2022,19(4):536-545.XIAN J C,CAI H Z,XIONG Y C,et al.Ground-based towed transient electromagnetic imaging method based on deep learning[J].Chinese Journal of Engineering Geophysics,2022,19(4):536-545.(in Chinese)

范涛，薛国强，李萍，等．瞬变电磁长短时记忆网络深度学习实时反演方法[J]．地球物理学报，2022,65(9):3650-3663.FAN T,XUE G Q,LI P,et al.TEM real-time inversion based on long-short term memory network[J].Chinese Journal of Geophysics,2022,65(9):3650-3663.(in Chinese)

ASIF M R,MAURYA P K,FOGED N,et al.Automated transient electromagnetic data processing for ground-based and airborne systems by a deep learning expert system[J].IEEE Transactions on Geoscience and Remote Sensing,2022,60:5919814.

燕帅，殷长春，苏扬，等．基于卷积神经网络的瞬变电磁数据快速成像算法研究[J]．地球物理学报，2023,66(10):4290-4300.YAN S,YIN C C,SU Y,et al.Fast imaging of transient electromagnetic data based on convolutional neural network[J].Chinese Journal of Geophysics,2023,66(10):4290-4300.(in Chinese)

程久龙，王慧杰，徐忠忠，等．基于全卷积神经网络的钻孔瞬变电磁法岩层富水性预测研究[J]．煤田地质与勘探，2023,51(1):289-297.CHENG J L,WANG H J,XU Z Z,et al.Research on aquifer water abundance evaluation by borehole transient electromagnetic method based on FCNN[J].Coal Geology&Exploration,2023,51(1):289-297.(in Chinese)

古瑶，解海军，周子鹏，等．基于Attention机制的CNN-BiLSTM瞬变电磁实时反演方法[J]．煤田地质与勘探，2023,51(10):134-143.GU Y,XIE H J,ZHOU Z P,et al.An attention mechanism-based CNN-BiLSTM real-time transient electromagnetic method[J].Coal Geology&Exploration,2023,51(10):134-143.(in Chinese)

王云宏．基于Transformer神经网络的瞬变电磁法数值模拟[J]．煤田地质与勘探，2024,52(11):151-159.WANG Y H.Transient electromagnetic numerical simulation based on the Transformer neural network[J].Coal Geology&Exploration,2024,52(11):151-159.(in Chinese)

KANG H,BANG M,SEOL S J,et al.Deep-learning-based airborne transient electromagnetic inversion providing the depth of investigation[J].Geophysics,2024,89(2):E31-E45.

HOLLAND J H.Adaptation in natural and artificial systems:An introductory analysis with applications to biology,control,and artificial intelligence[M].Cambridge:MIT Press,1992.

HINTON G E,SALAKHUTDINOV R R.Reducing the dimensionality of data with neural networks[J].Science,2006,313(5786):504-507.

JONG K A.An analysis of the behavior of a class of genetic adaptive systems[D].Ann Arbor:University of Michigan,1975.

SRINIVAS M,PATNAIK L M.Genetic algorithms:A survey[J].Computer,1994,27(6):17-26.

蒋文斌，彭晶，叶阁焰．深度学习自适应学习率算法研究[J]．华中科技大学学报(自然科学版)，2019,47(5):79-83.JIANG W B,PENG J,YE G Y.Research on adaptive learning rate algorithm in deep learning[J].Journal of Huazhong University of Science and Technology(Natural Science Edition),2019,47(5):79-83.(in Chinese)

MUNKHOLM M S,AUKEN E.Electromagnetic noise contamination on transient electromagnetic soundings in culturally disturbed environments[J].Journal of Environmental and Engineering Geophysics,1996,1(2):119-127.

齐彦福，智庆全，李貅，等．考虑关断时间的地面瞬变电磁三维带地形反演[J]．地球物理学报，2021,64(7):2566-2577.QI Y F,ZHI Q Q,LI X,et al.Three-dimensional ground TEM inversion over a topographic earth considering ramp time[J].Chinese Journal of Geophysics,2021,64(7):2566-2577.(in Chinese)

浏览量

175

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于MEVMD与GA-CNN+LSTM的NO_x浓度动态预测模型研究

基于RBF核的SVM核参数优化算法

Rh_n(n=3～57)团簇基态物性的遗传算法研究(英文)

配电网无功补偿容量及位置的优化研究

一种基于高阶互累计量的遗传盲反卷积算法