Using deep neural networks coupled with principal component analysis for ore production forecasting at open-pit mines

Ore production is usually affected by multiple influencing inputs at open-pit mines.Nevertheless,the complex nonlinear relationships between these inputs and ore production remain unclear.This becomes even more challenging when training data(e.g.truck haulage information and weather conditions)are massive.In machine learning(ML)algorithms,deep neural network(DNN)is a superior method for processing nonlinear and massive data by adjusting the amount of neurons and hidden layers.This study adopted DNN to forecast ore production using truck haulage information and weather conditions at open-pit mines as training data.Before the prediction models were built,principal component analysis(PCA)was employed to reduce the data dimensionality and eliminate the multicollinearity among highly correlated input variables.To verify the superiority of DNN,three ANNs containing only one hidden layer and six traditional ML models were established as benchmark models.The DNN model with multiple hidden layers performed better than the ANN models with a single hidden layer.The DNN model outperformed the extensively applied benchmark models in predicting ore production.This can provide engineers and researchers with an accurate method to forecast ore production,which helps make sound budgetary decisions and mine planning at open-pit mines.

Solution to the Horizontal Crossing Problem of Transportation System at Key Points in Open-Pit Mines

In order to reduce the horizontal crossing transportation problems between coal trucks and stripping trucks,large and small vehicles,and transport trucks and belt conveyors at key points of open pit mine in production,the separate transportation mode of underpass bridge and overpass steel trestle is proposed to optimize the open pit development transportation system,so as to solve the practical problems that the horizontal cross of transport vehicles causes vehicle blockage,affects production schedule and production safety.The results show that the horizontal crossing road can be changed into a separate type of overpass steel trestle,which can realize the classified transportation of large and small vehicles,reduce the traffic density,make vehicles with different functions go their own way,eliminate the hidden danger of traffic accidents,and improve the production efficiency.

Cut slope reinforcement technique in open-pit mines

The design and practice in supporting the cut slope of an open-pit mine wereintroduced, in which the high pressure grouting method was used in reinforcing the weak formation inthe slopes. Based on a detailed geological survey of the slope, a theoretical analysis was carriedout, and the design parameters were proposed, where the Tresca or Mohr-Coulomb yield criteria wasemployed. A patent technology, named 'Technology of high pressure and multiple grouting in differentlevels within a single hole', was employed in the construction. Anchor bars were also installed asgrouting proceeds. This method combines anchoring and grouting comprehensively and was foundsuccessful in practice.

Safe production model for small mines

Presented a 'safe production model' that can be adopted by small mine opera- tors to achieve their production targets safely and efficiently.The model consists of eight elements ranging from management commitment and leadership to safety account-ability and communication.The model is developed considering the mine operators' resource limitations and the workers' training needs.The study concludes with a summary of a sample survey that is conducted to validate the model and estimate a parameter for each mine and determine its position in the safe production scale.

Circle geometric constraint model for open-pit mine ore-matching and its applications

The circle geometric constraint model （CGCM） was put forward for resolving the open-pit mine ore-matching problems （OMOMP）. By adopting the approaches of graph theory, block model of blasted piles was abstracted into a set of nodes and directed edges, which were connected together with other nodes in the range of circle constraints, to describe the mining sequence. Also, the constructing method of CGCM was introduced in detail. The algorithm of CGCM has been realized in the DIM1NE system, and applied to a short-term （5 d） program calculation for ore-matching of a cement limestone mine in Hebei Province, China. The applications show that CGCM can well describe the mining sequence of ore blocks and its mining geometric constraints in the process of mining blasted piles. This model, which is applicable for resolving OMOMP under complicated geometric constraints with accurate results, provides effective ways to solve the problems of open-pit ore-matching.

Mercury-Free Gold Extraction Using Borax for Small-Scale Gold Miners

Small-scale gold mining is the largest anthropogenic contributor of mercury pollution on planet Earth. The miners grind gold ore together with mercury in order to recover the fine gold grains. The gold amalgam is burned whereby the mercury evaporates and gold is left behind. This processing extracts only a fraction of the gold in the ore and the released mercury causes serious global health and environmental problems. However, a mercury-free method described below has proved to extract up to twice as much gold without need of investment in new processing equipment and without longer processing time. The method has successfully been tested in the Philippines, Indonesia, Tanzania, Bolivia and Zimbabwe.

基于双曲嵌入的露天矿区暗光环境下道路多目标检测模型

露天矿环境特殊,道路场景复杂多变,在光照不足时会导致矿区道路多目标识别不清、定位不准,进而影响检测效果,给矿区无人矿用卡车的安全行驶带来严重安全隐患。目前的道路障碍物检测模型不能有效解决矿区暗光环境对模型检测效果的影响,同时对矿区小目标障碍物的识别也有较大误差,不适用于矿区特殊环境下障碍物的检测与识别。针对上述问题,提出了一种基于双曲嵌入的露天矿区暗光环境下多目标检测模型。首先,在模型的图像预处理阶段引入卷积神经网路Retinex-Net对暗图像进行增强,提高图像清晰度;然后,针对数据集中特征过多而无重点偏好的问题,在加强特征提取部分添加全局注意力机制,聚集3个维度上更关键的特征信息;最后,在检测模型预测阶段引入双曲全连接层,以减少特征丢失,并防止过拟合现象。实验结果表明:(1)基于双曲嵌入的露天矿区暗光环境下道路多目标检测模型不仅对露天矿区暗光环境下的大尺度目标具有较高的分类与定位精度,对矿用卡车及较远距离的小尺度目标即行人也可准确检测与定位,能够满足无人矿用卡车在矿区特殊环境下驾驶的安全需求。(2)模型的检测准确率达98.6%,检测速度为51.52帧/s,较SSD、YOLOv4、YOLOv5、YOLOx、YOLOv7分别提高20.31%,18.51%,10.53%,8.39%,13.24%,对于矿区道路上的行人、矿用卡车及挖机的检测精度达97%以上。

多层坚硬顶板特厚煤层综放工作面小煤柱护巷技术

为研究小煤柱护巷巷道的稳定性及其阻隔同层位邻近采空区灾害的特征,以大同矿区石炭系坚硬顶板特厚煤层开采为工程背景,采用理论分析、实验研究、数值模拟等研究方法,从小煤柱护巷巷道的稳定性和其是否具有阻隔同层位邻近采空区有害气体能力2个角度,综合分析小煤柱的合理尺寸及其在不同采动阶段渗透率的演化特征。针对同忻煤矿石炭系煤层8210工作面沿空掘巷小煤柱开采具体的开采条件,建立了双关键层采场内外应力场叠加的力学模型,推导出双关键层条件下沿空掘巷小煤柱合理尺寸的计算关系式,理论确定8210工作面小煤柱合理尺寸为6.0 m;应用DJG–Ⅱ型煤岩渗流测试装备,研究不同采动阶段小煤柱渗透率演化特征,试验确定在第3采动阶段小煤柱渗透率较初始渗透率增大了23倍,该阶段小煤柱基本失去了阻隔邻近采空区有害气体的能力。根据理论研究结果,现场选取6 m小煤柱进行工业性试验;根据试验研究结果,开采试验过程中对小煤柱进行了改性降透措施,在小煤柱表面及顶板距煤柱帮1500 mm范围内喷射厚层混凝土(100 mm)。实践表明:工作面回采过程中小煤柱护巷的回采巷道发生了一定的变形量,但是巷道变形在安全可控范围之内,巷道稳定,可实现安全回采;开采过程中8210工作面上隅角CH4气体浓度远低于邻近8305工作面采空区CH4气体浓度,表明小煤柱经过改性降透后具备了阻隔同层位邻近采空区有害气体的能力;开采实践也进一步验证了理论和试验研究成果的合理性和科学性。研究成果可以为类似条件下小煤柱护巷技术的推广应用提供参考。

山区采动地表异常形变识别与损害原因

现有开采损害鉴定的研究成果主要针对平原区域,缺少对高强度开采条件下山区地表建(构)筑物损坏的研究。以山西省冶头村建筑的异常损害为例,首先基于短基线集差分干涉测量技术(small baselines subset interferometric synthetic aperture radar,SBAS-InSAR),利用20景Sentinel-1A雷达影像,反演了目标区2016-11-14至2017-07-12期间地表形变的空间分布形态和时间演变过程。然后,结合现场调查的房屋损坏特征及地表裂缝数据,从开采沉陷理论和坡体稳定性分析两个角度进行开采损害鉴定的研究工作。研究表明:SBAS技术能够实现对采动地表异常形变的有效识别和提取,可为异常损害规律研究和原因揭示提供数据支撑;开采引起边坡稳定性降低,导致坡体蠕变滑移是地表建筑异常损害的原因。相关结论对于沉陷控制理论的发展完善,对开展山区井工开采房屋损害鉴定工作具有借鉴意义。

融合边界注意力的特征挖掘息肉小目标网络

从结肠图像中分割息肉小目标病变区域对于预防结直肠癌至关重要,它可以为结直肠癌的诊断提供有价值的信息。然而目前现有的方法存在2个局限性:一是不能稳健捕获全局上下文信息,二是未能充分挖掘细粒度细节特征信息。因此,提出融合边界注意力的特征挖掘息肉小目标网络(transformer feature boundary network,TFB-Net)。该网络主要包括3个核心模块:首先,采用Transformer辅助编码器建立长程依赖关系,补充全局信息;其次,设计特征挖掘模块进一步细化特征,学习到更好的特征;最后,使用边界反转注意力模块加强对边界语义空间的关注,提高区域辨别能力。在5个息肉小目标数据集上进行广泛实验,实验结果表明TFBNet具有优越的分割性能。

SBAS-InSAR技术融合CNN-LSTM模型的矿区开采沉陷监测与预测

针对传统矿区开采沉陷监测方法耗费人力财力和预测预警模型较少的问题,研究提出一种基于短基线集合成孔径雷达干涉测量(Small Baseline Subset-Interferometry Synthetic Aperture Radar,SBAS-InSAR)技术和卷积神经网络(Convolutional Neural Networks,CNN)与长短期记忆网络(Long Short-Term Memory,LSTM)相结合的矿区开采沉陷监测预测方法。首先,利用SBAS-InSAR技术对建新煤矿进行矿区开采沉陷监测,获取了该矿区的年平均沉降速率和累计沉降值。用GNSS监测数据与SBAS-InSAR结果进行对比验证,其拟合效果较好。其次,在此基础上利用CNN-LSTM模型预测后6期沉降数据,其结果与CNN和LSTM预测结果进行对比。研究显示,CNN-LSTM模型的平均绝对误差(S_(MAE))和均方根误差(S_(RMSE))比单一的CNN和LSTM分别至少降低了44.8%和40.6%,其决定系数均高于98%。最后,进一步预测前6期和中6期沉降数据,验证了CNN-LSTM预测模型在时间上的一致性。因此,SBAS-InSAR融合CNN-LSTM模型在类似矿山开采沉陷监测和预测中有较好的应用前景。

基于YOLOv7−SE的煤矿井下场景小目标检测方法

目前的小目标检测方法虽然提高了小目标检测效果,但针对的多为常规场景,而煤矿井下环境恶劣,在井下小目标检测过程中存在小目标特征信息提取困难的问题。针对上述问题,提出了一种基于YOLOv7−SE的煤矿井下场景小目标检测方法。首先,将模拟退火(SA)算法与k−means++聚类算法融合,通过优化YOLOv7模型中初始锚框值的估计,准确捕捉井下小目标;然后,在YOLOv7骨干网络中增加新的检测层得到井下小目标高分辨率特征图,减少大量煤尘对井下小目标特征表示的干扰;最后,在骨干网络中的聚合网络模块后引入双层注意力机制,强化井下小目标的特征表示。实验结果表明:①YOLOv7−SE网络模型训练后的损失函数值稳定在0.05附近,说明YOLOv7−SE网络模型参数设置合理。②基于YOLOv7−SE网络模型的安全帽检测平均精度(AP)较Faster R−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7分别提升了13.86%,25.3%,16.13%,12.71%,15.53%,11.59%,12.20%。基于YOLOv7−SE网络模型的自救器检测AP较Faster R−CNN,RetinaNet,CenterNet,FCOS,SSD,YOLOv5,YOLOv7分别提升了12.37%,20.16%,15.22%,8.35%,19.42%,9.64%,7.38%。YOLOv7−SE网络模型的每秒传输帧数(FPS)较Faster R−CNN,RetinaNe,CenterNet,FCOS,SSD,YOLOv5分别提升了42.56,44.43,31.74,39.84,22.74,23.34帧/s,较YOLOv7下降了9.36帧/s。说明YOLOv7−SE网络模型保证检测速度的同时,有效强化了YOLOv7−SE网络模型对井下小目标的特征提取能力。③在对安全帽和自救器的检测中,YOLOv7−SE网络模型有效改善了漏检和误检问题,提高了检测精度。

两次采动影响下小煤柱巷道切顶卸压围岩控制技术

为解决多次采动影响下留小煤柱巷道矿压显现剧烈、围岩变形量大的问题,提出了“切顶卸压+顶、帮锚索补强”联合控制技术。以贾家沟煤矿10115运输巷为工程背景,分析了切顶卸压位置、高度与巷道围岩垂直应力分布的关系,给出切顶卸压关键参数,并研究了留小煤柱巷道在切顶卸压后受两次采动影响下的矿压显现规律。结果表明:切顶卸压后,煤柱上的应力峰值随切顶深度的增加呈指数降低;10115运输巷在邻近工作面一次采动过程中巷道围岩变形依次呈现出无变形、缓慢变形、快速变形和围岩稳定的变化规律,本工作面二次采动过程中巷道围岩变形依次呈现出明显变形和剧烈变形的规律;巷道围岩在受二次采动影响时变形更加剧烈,巷道变形量为一次采动时的3.0~7.5倍。

Failure characteristics and its influencing factors of talus-derived rock mass during open-pit mining

The failure characteristic of talus-derived rock mass continues to challenge quantitative hazard assessments in open-pit mining. Physical model test was used to assess the failure modes and mechanisms on talus-derived rock mass. The different types of failure modes of the talus-derived rock mass were introduced and a possible failure mechanism relation between the failure zone and the structure of the talus-derived rock mass was also shown. The physical model test results indicate that the rainfall has significant influence on the stability and failure modes of talus-derived rock mass during open-pit mining. The development of the seepage area caused by rainfall initiates the localized failure in that particular area, and the initiation of localized instability is mainly induced by stress changes concentrated in the seepage area.

面向不确定性决策分析的数据获取模型与方法综述

不确定性决策是指在存在未知或难以准确预测结果情况下做出决策的过程。该文旨在研究在不确定性决策中,用4种数据获取方法来应对不确定性。这些方法包括:①使用小数据集;②利用开放数据集;③采用数据挖掘视角下的情感分析和遥感分析;④通过互联网问卷调查收集数据。该文首先回顾4种数据获取方法,并阐述各自的主要特点及发展历程。其次,阐述这4种数据获取方法在不确定性决策中的应用,并附有说明性示例。然后,讨论4种数据获取方法的局限性,并提出了未来研究的潜在趋势。最后,对全文进行总结。总体来说,该文为不确定性决策问题的研究提供了理论探讨和实践指导,以推动人工智能和决策科学领域的发展和创新,为决策者提供有效的数据支持,从而提高决策的科学性和准确性。

兰陵铁矿小口径定向钻进PDC钻头设计与应用

小口径定向钻进是解决高精度控矿勘探钻进的重要技术。针对兰陵铁矿区中硬岩石薄软互层小口径定向钻进施工碎岩PDC钻头广谱适应性问题,通过对钻头的冠部形状、刀翼布置、PDC切削齿镶焊角度、钻头水力结构及钻头胎体等几个方面的优化设计,研制了Ø98 mm定向钻进PDC钻头。经过中等硬度以上非均质地层3个定向钻探孔的应用验证,PDC钻头在抗冲击破坏、耐磨性、钻头寿命等方面取得了满意的效果,达到高效碎岩、连续稳定控向钻进的目的,为小口径定向钻进PDC钻头的设计及制造提供了一个有效解决方案。

针刺治疗非小细胞肺癌有效处方的选穴规律分析

【目的】分析针刺治疗非小细胞肺癌的选穴规律,为临床应用提供参考依据。【方法】检索针刺治疗非小细胞肺癌的临床研究、医家经验和学术思想以获取现代医家治疗非小细胞肺癌的针刺处方,基于穴位频次、组方规律、穴位聚类和核心组合等统计方法,剖析现代医家使用针刺治疗非小细胞肺癌的核心理论体系以及选穴规律。【结果】根据纳排标准最终纳入94个针刺处方。使用频次>20次的穴位有8个,分别是足三里、内关、肺俞、三阴交、气海、关元、合谷和中脘。高频穴位组合排前3位的分别为足三里-内关,足三里-三阴交和气海-足三里;关联分析显示,足三里-内关的关联强度最高,其次为足三里-三阴交、足三里-合谷。聚类分析显示:频次>10次的穴位可以分为三类。类别一包括:足三里、内关、三阴交、合谷;类别二包括关元、气海、中脘、血海、太冲;类别三由两大部分组成,其一是膻中、天突、丰隆、太渊,其二是肺俞、列缺、尺泽、中府、心俞、膏肓、风门。【结论】针刺治疗非小细胞肺癌的核心选穴包括4类:(1)健脾益气:足三里、三阴交、中脘和丰隆;(2)补益元气:关元、气海和膏肓;(3)理气宽胸:膻中、内关、天突和风门;(4)宣肺透气:肺俞、合谷、尺泽和列缺。补气、调气和透邪三大治法是晚期非小细胞肺癌针刺治疗的基础,以补为主,以疏为辅,疏补并用,是针刺治疗非小细胞肺癌的核心思路。

Comprehensive analysis of multiple machine learning techniques for rock slope failure prediction

In this study,twelve machine learning(ML)techniques are used to accurately estimate the safety factor of rock slopes(SFRS).The dataset used for developing these models consists of 344 rock slopes from various open-pit mines around Iran,evenly distributed between the training(80%)and testing(20%)datasets.The models are evaluated for accuracy using Janbu's limit equilibrium method(LEM)and commercial tool GeoStudio methods.Statistical assessment metrics show that the random forest model is the most accurate in estimating the SFRS(MSE=0.0182,R2=0.8319)and shows high agreement with the results from the LEM method.The results from the long-short-term memory(LSTM)model are the least accurate(MSE=0.037,R2=0.6618)of all the models tested.However,only the null space support vector regression(NuSVR)model performs accurately compared to the practice mode by altering the value of one parameter while maintaining the other parameters constant.It is suggested that this model would be the best one to use to calculate the SFRS.A graphical user interface for the proposed models is developed to further assist in the calculation of the SFRS for engineering difficulties.In this study,we attempt to bridge the gap between modern slope stability evaluation techniques and more conventional analysis methods.

基于改进YOLOv8n的采掘工作面小目标检测方法

为有效检测和识别煤矿井下采掘工作面人员是否佩戴安全防护装置,针对井下光照条件差、安全防护装备目标尺寸小且颜色与背景相似等情况,提出了一种基于改进YOLOv8n的采掘工作面小目标检测方法。在YOLOv8n骨干网络C2f模块中融合动态蛇形卷积(DSConv),构建C2f−DSConv模块,以提高模型提取多尺度特征的能力;在Neck层引入极化自注意力(PSA)机制,以减少信息损失,提高特征表达能力;在Head层增设1个专门针对小目标的检测头,形成4检测头结构,以扩大模型检测范围。实验结果表明,改进YOLOv8n模型对井下人员及其所佩戴安全帽、矿灯、口罩、自救器检测的平均精度分别为98.3%,95.8%,89.9%,87.2%,90.8%,平均精度均值为92.4%,优于Faster R−CNN,YOLOv5s,YOLOv7,YOLOv8n模型,且检测速度达208帧/s,满足煤矿井下目标检测精度和实时性要求。

基于潜在数据挖掘的小样本数据库对抗攻击防御算法

为了降低小样本数据库欺骗率,提升小样本数据库的攻击防御效果,设计了一种基于潜在数据挖掘的小样本数据库对抗攻击的防御算法(潜在数据挖掘的防御算法).采用改进的Apriori算法,通过频繁属性值集的工作过程获取准确的强关联规则优势,并从小样本数据库中挖掘潜在数据对抗攻击,同时优化候选集寻找频繁集的过程,然后利用关联分析检测对抗攻击,并通过可信度调度控制访问速率来防止产生恶意会话,实现小样本数据库对抗攻击防御.实验结果表明,潜在数据挖掘的防御算法可有效防御小样本数据库遭受的多种类型攻击,降低攻击产生的数据库欺骗率,保障小样本数据库服务器利用率的稳定性.