| In the exploration of lithium mines,many cores will be drilled,and the cores can accurately identify the mineralization and surrounding rock lithology to guide deep prospecting.In practical work,it is difficult for geologists to accurately distinguish the lithology of different rock layers by the naked eye.Therefore,it is necessary to carry out research on the lithology identification of cores.This paper is supported by the National Key R&D Project “Deep Exploration Technology Demonstration of Lithium Energy Metal Mine Base”(Project No: 2017YFC0602700).The core of the five-card drilling core of Jiajika is the research object,and the following research work is mainly carried out:1)Based on the principle of geochemistry,the geochemical characteristics of ore-bearing pegmatite,non-mineral pegmatite and surrounding rock in the Jiajika mine area were analyzed,and the components of major and trace elements in different lithologies were discussed.The difference provides a geochemical basis for distinguishing between ore-bearing pegmatites,non-mineral pegmatites,and surrounding rocks of different lithologies.2)Using the portable X-ray fluorescence(XRF)instrument as the tool,the X-ray fluorescence(peak area)characteristics of six main lithologies in the Jiajika mining area were studied,and the measured characteristic element X fluorescence information was established as different lithology.Discrimination basis,"Core XRF artificial neural network identification method" for identifying the lithology and ore-bearing properties of Jiajika drilling cores using artificial neural network model.Before the modeling of "Core XRF Artificial Neural Network Identification Method",the paper commissioned the Beijing Institute of Geology of the Nuclear Industry to collect 15 ore-bearing pegmatites,non-mineral pegmatites and surrounding rocks collected from the Jiajika area.The composition and content of ore samples were tested.The X-ray fluorescence spectrum(peak area)characteristics of these specimens were analyzed by IED-2000 T portable X-ray fluorescence instrument developed by Chengdu University of Technology.Based on the above work,An indicator element between each rock ore.Then,using the IED-2000 T portable X-ray fluorescence spectrometer,using the mineralization section 0.1-0.2m point spacing,the surrounding rock section 0.5-1.0m point distance one by one measured the methyl card mining area five drilling,a total of 920.5m core The main characteristic element of the X-ray fluorescence peak area.A total of 2,461 X-ray fluorescence data were obtained from the five-hole drilling XRF measurement,including 1245 YZK801 holes,380 YZK1201 holes,257 YZK3102 holes,346 YZK201 holes,and 233 YZK5102 holes.In the modeling of "core XRF artificial neural network identification method",based on the Jiajika ore and core XRF measurement data,the well core with geological catalogue data with accurate lithology is used as the standard,and the screening can be used as identification.Five XRF characteristic elements of different lithology and mineral layers.There are six kinds of lithology in the five cores of Jiajika : the andalusite schist,the spodumene pegmatite,the mica schist,the electric petrochemical mica schist,the non-mineral pegmatite and the cross stone schist.Based on the quantification of lithology at different depths of each borehole core,based on the difference between the components of different lithology major and trace elements and the XRF spectrum(peak area),artificial neural network modeling method is applied to different rocks.Sexual identification.Compared with the geological catalogue data of the borehole,the established core XRF artificial neural network identification method has a correct rate of 71.4% for the three-hole prediction core lithology identification,and the correct rate for the five-hole borehole core lithology recognition is 88.5 %. |
PDF Full Text Request