辽宁省北沟地区金矿地球化学特征及成矿预测

通过对北沟地区地球化学异常和已知矿体的研究,结合地质成矿环境,确定地球化学找矿指标,预测该区金矿找矿远景.     

岩溶塌陷灾害的岩溶地下水气压力监测技术及应用

文章系统、全面地介绍了岩溶地下水气压力监测技术及其应用成果。该技术在监测成孔、孔口的密封及监测频率等方面创新性地提出了相关的工艺方法,能真实、及时且充分地反映出岩溶管道裂隙系统中的地下水气压力变化特点,可为岩溶塌陷的机理研究、监测和预警提供科学依据。岩溶地下水气压力监测技术工艺简单、操作方便、成本低廉,经过20多年的改进,已成功应用于全国11个典型岩溶塌陷区,服务于高铁、水源地、油气管线、市政建设等潜在岩溶塌陷风险性评价、安全降深、监测预警等方面,取得了很好的效果。      

卷积神经网络及其在矿床找矿预测中的应用——以安徽省兆吉口铅锌矿床为例

大数据人工智能地质学刚刚起步,基于大数据智能算法的地质研究是非常有意义的探索性实验。利用大数据和机器学习解决矿产预测问题,有助于人们克服不能全面考虑地质变量的困难及评估当前模型在已有数据中的可靠性。元素地表分布特征量主要受原岩成分、成矿作用影响和地表过程的影响,它们携带某些指示矿体就位的信息,即矿体在地下空间就位时在地表的响应,且未在地表过程中消失。以往的地球化学勘查工作仅仅识别异常,但未能发现矿体在地表响应的成矿特征量。本文以安徽省兆吉口铅锌矿床为例,通过机器学习,利用卷积神经网络算法,不断挖掘元素Pb分布特征与矿体地下就位空间的耦合相关性。经过1000次训练后,可以得到准确率0. 93,损失率0. 28的卷积神经网络模型。这种神经网络模型就是矿体在地下就位时元素在地表分布的响应,可以用来进行矿产资源预测。应用该模型对未知区进行预测,结果显示第53号区域具有很大概率存在尚未发现的矿体。     

基于CART算法的三维成矿预测研究——以安徽白象山矿区为例

把地质大数据和人工智能技术引入矿产资源定量评价及成矿预测体系中,提高了海量地质数据的有效信息挖掘,弥补了传统方法的不足。本文基于白象山矿区基础地质资料和物化探成果资料,利用三维地质体建模技术和三维空间分析技术,量化三维控矿因素,建立了一种基于CART 算法的三维成矿预测模型。通过在白象山矿区的实验表明：该模型能较好的定位已知矿体,并且预测出在已知矿体北部、东部、东北部、西部、南部和东南部具有较高的成矿概率,可圈定找矿靶区。该模型将地质大数据应用于找矿勘探工作,具有纯数据驱动、预测精度高、预测结果可靠等优点。研究发现,该模型的预测效果与训练数据集的数量、矿控因素提取、决策树深度等有关。     

大数据环境下内蒙古浩布高地区铅锌多金属矿智能矿产预测研究

传统机器学习算法已广泛应用于矿产预测,但面对地质大数据的高维稀疏、不平衡小样本等特性仍缺乏有效处理和分析的方法,设计适合地质大数据特点的机器学习算法是智能矿产预测亟需解决的新问题。本文以内蒙古浩布高地区的铅锌多金属矿产预测为例,提出了一种面向地质大数据的半监督协同训练矿产预测模型。首先对研究区地质找矿信息和地球化学异常信息进行定量分析,提取断裂构造、二叠系地层、燕山期侵入岩、地层与岩体接触带、围岩蚀变及Pb、Zn、Sn、Cu地球化学异常共9种找矿因子。然后利用递归特征消除法优选找矿因子组合,不包括Sn异常在内的8个找矿因子组合被选为最优组合。最后,利用支持向量机和随机森林算法作为基分类器进行半监督协同训练矿产预测,绘制成矿概率分布图。ROC曲线和预测度曲线分析结果表明,半监督协同训练模型的AUC值和预测效率都高于随机森林和支持向量机模型。研究结果也为大数据环境下的智能矿产预测提供了一种新的思路。     

音频大地电磁测深在齐大山铁矿王家堡子采区深部找矿中的应用

鞍山齐大山铁矿王家堡子采区产出鞍山式沉积变质型铁矿,铁矿石多以磁铁贫矿为主,局部产出富铁矿,目前钻探工程控制标高为-600 m,采区深部的资源潜力是急待解决的问题,为此对采区进行音频大地电磁测量和研究工作。齐大山铁矿的电性变化复杂,通过先期的实验剖面确定了矿区铁矿体的3种不同电阻率特征,然后对王家堡子采区的3760线、4050线、4500线进行音频大地电磁测量、数据处理和断面反演分析。推测3760线的低阻由磁铁贫矿引起,局部高阻为假象赤铁贫矿引起;推测4050线和4500线浅部的低阻由磁铁贫矿引起,深部的高阻为假象赤铁贫矿及磁铁贫矿引起。同时预测3760线、4050线和4500线西侧深部均有低电阻率显示,反映出隐伏铁矿床的存在特征,推测在-1 000 m标高以下仍存在有隐伏的富矿体或板状磁铁贫矿。经过钻探验证,获得了预期的找矿效果。     

Progress in the Application of Ground Gravity Observation Data in Earthquake Prediction

The mechanism of earthquake inoculation and the process of earthquake occurrence are very complicated. Additionally, earthquakes do not happen very often, and we lack enough cognition to the earth’s interior structure, activity regularity and other key elements. As a result, research progress about the theory of earthquake precursors has been greatly restricted. Ground gravity observation has become one of the main ways to study earthquake precursor information in many countries and regions. This paper briefly summarized the surface gravity observation technology and observation network in China: the surface gravity measurement instrument developed from Huygens physical pendulum in seventeenth Century to today’s high-precision absolute gravimeter, and its accuracy reached to ±1×10^-8 m/s². China has successively established the National Gravity Network, Digital Earthquake Observation Network of China,the Crustal Movement Observation Network of China Ⅰ and the Crustal Movement Observation Network of China, to provide a public platform for monitoring non tidal gravity change, seismic gravity and tectonic movement. The use of specific examples illustrated the role of gravity observation data in earthquake prediction. The gravity observation data of ground gravity can be used to capture the information of gravity change in the process of strong earthquake inoculation, and to provide an important basis for the long-term prediction of strong earthquakes. The temporal and spatial variation characteristics of the regional gravity field and its relation to strong earthquakes were analyzed: Before the earthquake whose magnitude is higher than M_S 5, generally there will be a large amplitude and range of gravity anomaly zones. Strong earthquakes occur mainly in areas where the gravity field changes violently. The dynamic change images of gravity field can clearly reflect the precursory information of large earthquakes during the inoculation and occurrence. Finally, the existing problems of surface gravity technology in earthquake precursor observation were put forward and the use of gravity measurement data in earthquake prediction research was prospected.     

川西裂缝性致密碎屑岩气藏高渗区成因机理及预测

川西深层上三叠统须家河组二段致密碎屑岩气藏非均质性极强,裂缝网络与相对优质储层叠加形成的高渗区分布预测是钻井获得高产稳产的技术关键。通过地表露头调查、钻井资料统计分析及地震地质综合研究,剖析了高渗区地质成因机理,认为在砂岩及岩性相变带、相对薄层及地层厚度突变带、优质储层发育区、强构造褶皱变形区、断裂附近等易于形成裂缝及高渗区。通过高渗区储层测井及波阻抗特征、地震响应特征对比分析、模型正演等方法,建立了“杂乱弱反射”地震暗点识别模式,利用吸收处理、像素成像、三维可视化等地震技术手段,预测了高渗区分布,解释了各高渗区地质成因,钻井验证了预测结果。     

Settlement prediction by spatial-temporal random process using Asaoka's method

A methodology was presented for observation-based settlement prediction with consideration of the spatial correlation structure of soil. The spatial correlation is introduced among the settlement model parameters and the settlements at various points are spatially correlated through these geotechnical parameters, which naturally describe the phenomenon. The method is based on Bayesian estimation by considering both prior information, including spatial correlation and observed settlement, to search for the best estimates of the parameters at any arbitrary points on the ground. Within the Bayesian framework, the optimised selection of auto-correlation distance by Akaike's Bayesian Information Criterion (ABIC) is also proposed. The application of the proposed approach in consolidation settlement prediction using Asaoka's method is presented in this paper. Several case studies were carried out using simulated settlement data to investigate the performance the proposed approach. It is concluded that the accuracy of the settlement prediction can be improved by taking into account the spatial correlation structure and the proposed approach gives the rational prediction of the settlement at any location at any time with quantified uncertainty.     

Modelling real earthquake activity with reverse engineering based on evolutionary computation methods

Hazard assessment of dangerous natural phenomena is critical because of their evident results concerning loss of human life and property, especially in dense populated areas. Earthquakes are probably the most devastating phenomenon since their immediate and long-term consequences are severe. This study is focused on the earthquake data analysis in different regions of Greece, characterised by different seismicity levels. In specific, a novel model is proposed based on evolutionary computation methods, such as symbolic regression by genetic programming and genetic algorithms in order to elucidate preliminary hidden mathematical relations and patterns found in the seismological signals under study. Furthermore, the model is calibrated using reverse engineering and closes the loop from the data collection to initial hypothesis. In this way, the model formation is achieved. The presented simulation results qualitatively and quantitatively reveal some of the fundamental characteristics of each studied geographical region located in Greece that stem from its geodynamic properties.