基于深度学习的镜下矿石矿物的智能识别实验研究

矿石矿物鉴定的智能化是智能地质学和智能矿床学的基础技术之一。计算机视觉技术和深度学习理论使矿石矿物鉴定的智能化成为可能。本研究基于深度学习系统Tensor Flow,以吉林夹皮沟金矿和河北石湖金矿的黄铁矿、黄铜矿、方铅矿、闪锌矿等硫化物矿物为例,设计有针对性的Unet卷积神经网络模型,有效自动提取矿相显微镜下矿石矿物的深层特征信息,实现镜下矿石矿物智能识别与分类。实验显示,模型在训练过程中,随着训练次数的增加,模型精度在不断增大,损失函数不断减小;经过3000个批处理之后,模型精度和损失函数基本趋于稳定。训练出的模型对测试集中的显微镜镜下矿石矿物照片的识别成功率均高于90%,说明实验所建立的模型,具有很好的图像特征提取能力,能完成镜下矿石矿物智能识别的任务。     

石英Ti/Ge-P:基于机器学习的矿床类型判别新图解SCIEI北大核心CSCD

石英的微量元素记录了石英生长的物理化学条件。通过微量元素对石英原岩进行分类的研究历史已久,经典工作是在以微量元素为坐标轴的图解上绘制各类型石英的分布范围,以区分石英类型。经典图解包括Rusk(2012)提出用于区分三种矿床类型石英的Al-Ti二元图解,和Schr9n et al.(1988)提出的用于判别不同岩浆岩类型石英的Ti-Al-Ge三元图解。越来越多的研究表明,上述图解不能满足对更多石英类型进行分类的需求,同时也出现与部分已知产状类型的石英微量元素判别相矛盾的情况。随着石英原位微区测试方法的成熟,高精度石英微量元素数据逐渐丰富为系统开展机器学习提供了大数据基础,为石英微量元素研究提供了新的角度和可能性。本研究运用机器学习分类方法对石英微量元素进行精确数学分析,提出Ti/Ge-P图解为石英成因研究提出新的地球化学指标。本文同时测试了六种经典机器学习分类算法,提高Ti/Ge-P图解在石英成因分类研究上的精度。此Ti/Ge-P图解适用于多种矿床研究,包括但不局限于斑岩型矿床、矽卡岩型矿床、浅成低温热液型矿床、卡林型矿床以及造山型矿床中的石英。这项工作是大数据技术与机器学习技术在地球化学研究中的积极探索。     

Effects of crude oil contamination on geotechnical properties of clayey and sandy soils

The southern coastal plain of Iran at the Persian Gulf encounters oil pollution due to the historical oil exploitation, related tanker navigations and accidents, and petrochemical industrial expansions in the recent years. Therefore, it is important to investigate the geochemical properties of oil-contaminated coastal soils and sediments for engineering and environmental purposes. Here, an extensive laboratory testing program was carried out to determine the effects of crude oil contamination on some of the geotechnical properties of clayey and sandy soils such as CL, SM and SP sampled from the coastal soils from this area. The testing included basic properties, Atterberg limits, compaction, direct shear, uniaxial compression and permeability tests on clean and contaminated soil samples at the same densities. The contaminated samples were prepared by mixing the soils with crude oil in the amount of 2%, 4%, 8%, 12%, and 16% by dry weight. The results indicated a decrease in strength, permeability, maximum dry density, optimum water content and Atterberg limits. Knowledge of these effects of oil contamination is important in coastal engineering and environmental remediation activities of the studied coastal plain.     

原油物性演化规律新发现及其对页岩油勘探的启示

针对烃源岩生成的原油物性随热演化如何变化的问题,本文基于对准噶尔盆地玛湖凹陷下二叠统风城组和吉木萨尔凹陷中二叠统芦草沟组下段源储紧邻自生自储页岩油物性、地化特征、成藏特征及原油物性与烃源岩热演化关系的分析,综合烃源岩生烃热模拟实验结果与国内外相关文献资料,首次确认了咸化湖相烃源岩生成的原油物性与烃源岩热演化程度之间的变化规律。认为页岩油源岩生成的原油密度和黏度具有随热演化程度增强先增加而后降低的规律,其中生油高峰附近生成的原油非烃相对含量、密度和黏度最高。该认识不仅是对已有的石油地质学中原油物性随热演化规律认识的进一步厘定和修正,而且对页岩油甜点段、甜点区的选择以及页岩油原位转化等都有极为重要和现实的指导作用。     

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。     

原油可采储量预测新方法探讨

原油可采储量是制定油田开发规划的基础，是评价油田开发效果、编制调整方案的依据．原油可采储量预测方法主要有静态法、童宪章图版法、水驱特征曲线法、递减曲线法和增长曲线法．作提出了原油可采储量预测的新方法，改进流管法和随机模型法．它们是作多年研究的体会，其适用性正在大港油田原油可采储量预测工作中进行检验，已取得较好的效果．     

A graph deep learning method for landslide displacement prediction based on global navigation satellite system positioning

The accurate prediction of displacement is crucial for landslide deformation monitoring and early warning. This study focuses on a landslide in Wenzhou Belt Highway and proposes a novel multivariate landslide displacement prediction method that relies on graph deep learning and Global Navigation Satellite System (GNSS) positioning. First model the graph structure of the monitoring system based on the engineering positions of the GNSS monitoring points and build the adjacent matrix of graph nodes. Then construct the historical and predicted time series feature matrixes using the processed temporal data including GNSS displacement, rainfall, groundwater table and soil moisture content and the graph structure. Last introduce the state-of-the-art graph deep learning GTS (Graph for Time Series) model to improve the accuracy and reliability of landslide displacement prediction which utilizes the temporal-spatial dependency of the monitoring system. This approach outperforms previous studies that only learned temporal features from a single monitoring point and maximally weighs the prediction performance and the priori graph of the monitoring system. The proposed method performs better than SVM, XGBoost, LSTM and DCRNN models in terms of RMSE (1.35 mm), MAE (1.14 mm) and MAPE (0.25) evaluation metrics, which is provided to be effective in future landslide failure early warning.     

Effect of secondary oil migration distance on composition of acidic NSO compounds in crude oils determined by negative-ion electrospray Fourier transform ion cyclotron resonance mass spectrometry

The acidic and neutral NSO compounds in a series of Duvernay-sourced oils in Canada, which are believed to have migrated extensively over relatively long distances such as along the Rimbey-Meadowbrook reef trends, were characterized by negative-ion electrospray (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). Heteroatomic compounds were characterized according to their class (number of nitrogen, oxygen and sulfur heteroatoms), degree of aromaticity [rings plus double bonds (DBE)] and carbon number distribution. The N₁, N₁O₁, N₁O₂, N₁S₁, O₁ and O₂ classes were identified in Duvernay-sourced oils. With increasing migration distance, the relative abundance of O₂, N₁O₁ and N₁O₂ showed a significant decrease, while the O₁ class increased from < 10% to nearly 30% of the total. With increasing migration along the Rimbey-Meadowbrook reef trend, pyrrolic nitrogen compounds (N₁ class) shows an enrichment of alkylcarbazoles (DBE = 9) relative to alkylbenzocarbazoles (DBE = 12), and of higher homologous relative to the lower homologous. O₁ compounds show a relative enrichment of those with low DBE values. Additionally, the N₁O₁ and N₁S₁ compounds show a relative enrichment of those with high DBE values, and of higher homologues compared to the lower homologues, indicating great potential for developing new migration indices.     

A prediction model for important mineral resources in China

Prediction models for mineral resources provide an analytical foundation and method to express the results of resource evaluations. The project “China National Mineral Resources Assessment Initiative” was conducted during 2006–2013, with the aim to determine the location, quantity, and quality of 25 important mineral resources occurring at depths of <1 km. There are currently 80 integrated prediction models on the scale of III–level metallogenic belts in use across China. The Huangshaping Pb–Zn polymetallic deposit, Hunan province, China, is used as a case study to establish methods and processes for developing a mineral resource prediction model that would be used for exploration targeting. The construction of prediction models requires the development of a classification scheme for the proposed prediction method appropriate for the prediction area. An initial metallogenic model is quantitatively transformed to a prospecting model, and then a prediction model. The incorporation of additional methodology, analysis of a comprehensive geological database, and correlation of asymmetric information between the well–explored typical deposit area and regional prediction area, yield an integrated prediction model. This paper also discusses the prediction modeling theory, and presents 12 models used for mineral assessments.     

深度学习在地质学上的应用

自Hinton等使用基于卷积神经网络的深度学习模型赢得Image Net分类比赛以来,深度学习的研究席卷了各个行业。通过介绍深度学习的历史,探索国内地质行业中深度学习模型的使用情况,并介绍深度学习的基础概念(如神经元、神经网络、监督学习和无监督学习等)以及深度学习基础模型中的2个重要网络:深度信念网络(DBN)和卷积神经网络(CNN)。在此基础上,类比深度学习在医学等相关领域的应用,提出了深度学习在地质上的几点应用:利用深度学习在计算机视觉上表现出的强大能力,可以对遥感图像进行聚类、对岩石样品图像进行分类、对岩石薄片数据进行描述;利用深度学习对原始数据表现出的强大识别能力,处理地质异常数据,从而确定成矿靶区的可能位置;利用深度学习的特点,对地震前的声信号数据进行处理,从而判断出地震发生前的剩余时间。     

Advanced prediction of tunnel boring machine performance based on big data

Predicting the performance of a tunneling boring machine is vitally important to avoid any possible accidents during tunneling boring.The prediction is not straightforward due to the uncertain geological conditions and the complex rock-machine interactions.Based on the big data obtained from the 72.1 km long tunnel in the Yin-Song Diversion Project in China,this study developed a machine learning model to predict the TBM performance in a real-time manner.The total thrust and the cutterhead torque during a stable period in a boring cycle was predicted in advance by using the machine-returned parameters in the rising period.A long short-term memory model was developed and its accuracy was evaluated.The results show that the variation in the total thrust and cutterhead torque with various geological conditions can be well reflected by the proposed model.This real-time predication shows superior performance than the classical theoretical model in which only a single value can be obtained based on the single measurement of the rock properties.To improve the accuracy of the model a filtering process was proposed.Results indicate that filtering the unnecessary parameters can enhance both the accuracy and the computational efficiency.Finally,the data deficiency was discussed by assuming a parameter was missing.It is found that the missing of a key parameter can significantly reduce the accuracy of the model,while the supplement of a parameter that highly-correlated with the missing one can improve the prediction.     

基于机器学习的三维矿产定量预测——以四川拉拉铜矿为例

在大数据蓬勃发展的时代背景下,矿产资源定量预测作为地质大数据的核心部分,其综合分析挖掘多元信息的基本思路与大数据的理念不谋而合。以四川拉拉铜矿为例,开展基于机器学习的三维矿产资源定量预测。通过建立三维地质模型,提取成矿有利信息,构建研究区定量预测模型;基于"立方块预测模型"找矿方法,采用机器学习随机森林算法,计算出研究区成矿概率分布,以此圈定出5个找矿远景区。结果表明,随机森林具有更高的预测准确度与稳定性,且能够对控矿要素重要性做出定量评价。该研究成功地将机器学习应用于三维矿产定量预测,为今后的矿产资源预测评价做出了积极的探索。     

Developing a 3D elastoplastic constitutive model for soils: A new approach based on characteristic stress

This paper presents a new approach for the development of an elastoplastic constitutive model to predict the strength and deformation behaviour of soils under general stress conditions. The proposed approach was based on characteristic stress, which considers the effect of the intermediate principal stress on the material strength. Referring to the Cam-clay model, the shear dilatancy equation, plastic potential function and hardening parameter for the developed model were all derived using the characteristic stress. The model predictions indicated that the established model could quantitatively reproduce the negative dilatancy behaviour, positive dilatancy behaviour, and three-dimensional strength properties of soils.     

中国不同类型盆地油气资源丰度统计特征及预测模型

油气资源丰度是油气资源评价的一项关键参数,取准资源丰度对于客观评价资源潜力十分重要。本文立足于类比刻度区进行统计分析,探索建立资源丰度预测模型的思路并开展研究;统计表明我国主要含油气盆地的常规油气资源普遍为中低丰度,并以特低丰度和低丰度资源为主,仅局部优越区块为高丰度。据此筛选刻度区数据,并按照裂谷、克拉通、前陆3种基本盆地类型分析常规油气资源丰度与相关地质参数的关系。不同类型盆地的油气资源丰度均与烃源岩生烃强度、圈闭面积系数、区域不整合次数3项主控地质参数具有显著的相关性,但不同类型盆地油气资源丰度与3项主控地质参数之间关系的变化趋势存在差异,拟合的单参数地质模型能够较好描述3项主控地质参数单独与不同类型盆地油气资源丰度的定量关系。由此建立了基于烃源岩生烃强度、圈闭面积系数、区域不整合次数3项主控地质参数的油气资源丰度预测模型,能够快速、准确求取不同类型盆地的常规油气资源丰度。预测模型在实际应用中效果较为理想,具有易操作、易推广的优点,实用价值较高。     

A novel index to evaluate discretization methods:A case study of flood susceptibility assessment based on random forest

The selection of a suitable discretization method(DM)to discretize spatially continuous variables(SCVs)is critical in ML-based natural hazard susceptibility assessment.However,few studies start to consider the influence due to the selected DMs and how to efficiently select a suitable DM for each SCV.These issues were well addressed in this study.The information loss rate(ILR),an index based on the informa-tion entropy,seems can be used to select optimal DM for each SCV.However,the ILR fails to show the actual influence of discretization because such index only considers the total amount of information of the discretized variables departing from the original SCV.Facing this issue,we propose an index,infor-mation change rate(ICR),that focuses on the changed amount of information due to the discretization based on each cell,enabling the identification of the optimal DM.We develop a case study with Random Forest(training/testing ratio of 7:3)to assess flood susceptibility in Wanan County,China.The area under the curve-based and susceptibility maps-based approaches were presented to compare the ILR and ICR.The results show the ICR-based optimal DMs are more rational than the ILR-based ones in both cases.Moreover,we observed the ILR values are unnaturally small(＜1％),whereas the ICR values are obviously more in line with general recognition(usually 10％-30％).The above results all demonstrate the superiority of the ICR.We consider this study fills up the existing research gaps,improving the ML-based natural hazard susceptibility assessments.     

机器学习在岩矿地球化学研究中的应用——综述与思考

岩石与矿物的地球化学成分数据具有高维度特征。传统的岩矿地球化学成分研究主要采用二元/三元图解判别法,准确率不高,在数理统计方法上有欠缺。机器学习方法非常适用于对大样本高维度的岩矿成分数据进行数理统计处理。笔者等在介绍机器学习常见算法基本原理的基础上,总结近5年来国内外学者将机器学习方法应用于岩石矿物成分数据研究的实例,包括：(1)根据矿物成分溯源其母岩(源岩)、判别矿床类型,(2)新生代火山岩溯源,(3)判别变质岩原岩,(4)依据岩浆岩成分判别大地构造环境等。已有的研究实例显示,机器学习方法的准确度明显优于传统的低维度判别法。机器学习本质是分析大样本数据的高维度变量之间的相关、归类等多元统计问题。推广机器学习的应用需要建设开放获取(Open Access)的矿物、岩石成分数据库,同时全面实施开放研究(Open Research)的发表策略。     

机器学习在岩矿地球化学研究中的应用——综述与思考

岩石与矿物的地球化学成分数据具有高维度特征。传统的岩矿地球化学成分研究主要采用二元/三元图解判别法,准确率不高,在数理统计方法上有欠缺。机器学习方法非常适用于对大样本高维度的岩矿成分数据进行数理统计处理。本文在介绍机器学习常见算法基本原理的基础上,总结近5年来国内外学者将机器学习方法应用于岩石矿物成分数据研究的实例,包括：① 根据矿物成分溯源其母岩（源岩）、判别矿床类型,② 新生代火山岩溯源,③ 判别变质岩原岩,④ 依据岩浆岩成分判别大地构造环境等。已有的研究实例显示,机器学习方法的准确度明显优于传统的低维度判别法。机器学习本质是分析大样本数据的高维度变量之间的相关、归类等多元统计问题。推广机器学习的应用需要建设开放获取（Open Access）的矿物、岩石成分数据库,同时全面实施开放研究（Open Research）的发表策略。     

基于卷积神经网络的福建省区域滑坡灾害预警模型

福建省滑坡灾害频发,开展区域尺度上的滑坡灾害预警是防灾减灾的重要手段,但由于滑坡成灾机理复杂,传统的区域滑坡预警方法存在精度不足等问题。深度学习是指通过构建神经网络模型进行特征的提取、抽象、表示与学习的技术,是机器学习的一种。卷积神经网络作为一种经典的深度学习算法,具有比传统机器学习更强大的分类能力与表征能力。文章以福建省为研究区,将卷积神经网络引入滑坡灾害预警领域,构建福建省区域滑坡预警模型,过程及结果如下：（1）采用SMOTE优化算法对2010—2018年福建省滑坡灾害样本库进行优化,扩充正样本的个数,将正负样本比例从1∶3.4扩充到1∶2,样本总量达到18040个;（2）构建卷积神经网络模型结构,模型结构包括一个输入层、两个卷积层、两个最大池化层和一个全连接层以及一个输出层;（3）使用卷积神经网络对优化后的样本（2010—2018年样本的80%作为训练集）进行训练,并用贝叶斯优化算法优化模型超参数,得到福建省区域滑坡预警模型;（4）以2010—2018年样本的20%作为测试集对模型进行测试,采用混淆矩阵、ROC曲线进行模型测试,结果显示模型准确度为0.96～0.97,AUC值达到0.977,模型精度与泛化能力良好;（5）以2019年汛期滑坡灾害实况作为正样本,通过时空采样的方法采集负样本,构建2019年区域滑坡样本校验集（样本数603个）,对模型进行进一步实况校验,采用混淆矩阵、ROC曲线进行模型校验,结果显示模型准确度为0.75～0.85,AUC值为0.852。虽然仅用了2019年汛期的滑坡实况样本进行校验,但也达到较好的效果。将卷积神经网络算法应用到区域滑坡预警中,为建立区域滑坡预警模型提供了一种新的途径,初步校验表明,模型效果良好,今后将在福建省对模型进行进一步的应用与校验。     

A Bayesian network approach for predicting seismic liquefaction based on interpretive structural modeling

The Bayesian network (BN) is a type of graphical network based on probabilistic inference that has been gradually applied to assessment of seismic liquefaction potential. However, how to construct a robust BN remains underexplored in this field. This paper aims to present an efficient hybrid approach combining domain knowledge and data to construct a BN that facilitates the integration of multiple factors and the quantification of uncertainties within a network model to assess seismic liquefaction. Initially, only using given domain knowledge, a naive network model can be constructed using interpretive structural modeling. Thereafter, some effective information about the naive model is provided to construct a robust model using structural learning of BN from historic data. Finally, the returning predictive results and the predictive results are compared to other methods including non-probabilistic and probabilistic models for seismic liquefaction using the metrics of the overall accuracy, the area under the curve of receiver operating characteristic, prediction, recall and F₁ score. The methodology proposed in this paper achieved better performance, and we discussed the power and value of the proposed approach at the end of this paper, which suggest that BN is a good alternative tool for seismic liquefaction prediction.