摘要
在砂岩型铀矿勘探中,开展音频大地电磁法(Audio Magnetotelluric Method,AMT)测深能够有效地获得深部砂体发育范围、基底隆凹形态等与铀成矿相关的地质信息。其测量成像结果包含的地质意义往往需要通过主观地质推断后才能为后续地质勘探工作提供支撑,解译结果严重依赖解译人员的地质认知和专业经验。与此同时,随着三维AMT测量技术的发展,人工地质解译的效率较为低下。文章以砂岩型铀矿成矿环境为基础,采用随机电阻率模型构建AMT图像样本库,并建立U-net网络开展砂体、基底等岩性识别训练,从而实现智能推断反演电阻率体中的砂体和基底。通过与人工解译结果对比,智能解译结果匹配程度较高,解译时间大大缩短,表明该方法在砂岩型铀矿勘探中具有较好的应用前景。
AMT(Audio Magnetotelluric)is widely used for obtaining geological condition related to sandstone-type uranium deposits,such as the range of buried sand body and the top boundary of basement rock.However,these geological condition are hard to interpret via measured sections without geological deduction,which relies heavily on the experience and cognition of the interpreter.On the other hand,with the development of 3D technology,artificial geological interpretation shows low efficiency and reliability.In this paper,a deep learning model constructed using U-net was used for the geological interpretation of AMT data in the Naren-Yihegaole area.To train the model,a training dataset was built based on the simulated data from random simulated models.In the prediction stage,sand bodies and basement rock were delineated from the inversion resistivity images.The comparison between two interpretations,one by deep learning method,showed high consistency with the artificial one,but with better time-saving.Therefore,the deep learning based technology is more effective than the traditional way.
作者
汪硕
周渊凯
马国财
胡跃彬
胡渤
WANG Shuo;ZHOU Yuankai;MA Guocai;HU Yuebin;HU Bo(Beijing Research Institute of Uranium Geology,Beijing 100029,China;Qinghai Provincial Non-ferrous Metal Geological and Minerals Exploration Bureau,Xining,Qinghai 810001,China)
出处
《铀矿地质》
CAS
CSCD
2024年第4期803-808,共6页
Uranium Geology
基金
中核集团青年英才项目(编号:物QNYC2020-1),中核集团集中研发项目(编号:物SDEQ02)联合资助。
关键词
砂岩型铀矿
AMT
深度学习
智能地质解译
sandstone-type uranium
AMT
deep learning
artificial intelligent geological interpretation
