Ore-controlling Regularities of Thrust-fold structures and features of Tectono-geochemical Anomalies at the Xiaozhuqing Exploration Area in the Huize Zn-Pb District

1 Introduction The huize Zn-Pb ore district in Yunnan province is locatedinthecentralsouthernofthe Sichuan—Yunnan—GuizhouPb-ZnPoly-metallic Mineralization Area in the southwestern margin of the Yangtze Block,and is strictly controlled by fault structures.It has developed to one of the famous production bases of lead&zinc and germanium in China.

The Characteristics of Ore-Controlling Structure in Baoshan Cu-Pb-Zn Polymetallic Deposit,Hunan Province

The Baoshan Cu-Pb-Zn polymetallic deposit is lied in the central Nanling mineralization zone,and belongs to the junction area of the Chenzhou-Linwu fault zone and the Leiyang-Linwu fault zone.It is a significant part of Nanling polymetallic deposit belt.The outcropping stratas consist of upper Devonian Shetianqiao,Xikuangshan Formation,Lower Carboniferous Menggong’ao,Shidengzi,Ceshui,and Zimenqiao Formation.Igneous rocks in the Baoshan ore area mainly comprise granodiorite porphyry.Furthermore,the radio isotopic age ranges from 123 Ma to 183 Ma,belonging to the early to middle Yanshanian.

Analysis of Ore-controlling Structure in the Qifengcha-Detiangou Gold Deposit,Huairou County,Beijing

The Qifengcha-Detiangou gold deposit is a medium-sized deposit recently found in Huairou County, Beijing. It belongs to the altered mylonite type with superimposed quartz vein type and is related to the early Yanshanian magmatic activity. Characterized by multiperiodic activity, the NE-trending Qifengcha fault is a regional ore-controlling structure in the area, and gold mineralization develops only in its southeastern part. Meanwhile, gold mineralization is controlled by the Yunmengshan metamorphic core complex. The nearly N-S- and E-W-trending low-angle detachment faults, reformed by the Qifengcha fault in the northwestern part of the core complex, are the main ore-bearing faults. All discovered gold deposits are located within an area 1.5–4.0 km away from the boundary of the upwelling centre. The N-S- (NNE-) and E-W-trending ore-bearing faults are ductile-brittle structural zones developing in shallow positions and subjected mainly to compressive deformation. The structural ore-controlling effects are as follows. (1) The attitude, shape, and distribution of gold orebodies are controlled by faults. (2) There is a negative correlation between the gold abundance and the magnetic anisotropy (P) of the altered mylonite samples from the deposit, which shows that the gold mineralization is later than the structural deformation. (3) Quartz vein type mineralization is superimposed on altered mylonite type mineralization. (4) In mineralized mylonite, the stronger the ductile shear deformation, the easier the late-stage gold mineralization to occur and the higher the gold abundance. The richest gold mineralization occurs only around the centre of the fault subjected to the strongest deformation.

Analysis of Ore-Controlling Structures of the Xincheng-Hexi Gold Deposit,Shandong Province,China

Based on quantitative and semi-quantitative mathematical and mechanical analysis of the shape, motion, structural factors, stress field and deformation field of the ore-hosting faults in the Xincheng-Hexi gold deposit, the ore-controlling features of faults and mineralization mechanism are discussed. It is concluded that the mineralization is controlled by the main faults, subsidiary fractures, joint density, mechanical features and deformation of the faults. The ore bodies are mainly located in the lower part of the convex crest and upper part of the concave trough of the main undulating fault surface. Mineralization is positively correlated to the development of subsidiary fractures and joints, which correspond to zones of low internal stress and high body strain and shear strain. They are favourable positions for mineralization and alteration.

GEOCHEMICAL REASERCH ON ORE-CONTROLLING STRUCTURE IN LIAODONG LUOQUANBEI-BAIYUN GOLD MINERAL BELT

In recent2 0 years,because of the finding ofmany large- superlargegold deposits,we re- new the theories for prospecting and gain many things and conceptions.The current geo- chemical and gold- forming theories underline the importance of the early submarine volcano- sedimentation,metamorphic differention,sedimentation of terrigenous clastics,thermal spring and it’s sedimentation,syngenesis process and other hypergene supplying the source for metallogenic materials.According to the study for source bed(rock) and depsitional for- mation of gold,we find that gold will be gradually enriched and mineralized in source bed (rock) ,because of variousgeologicprocesses,such as regional metamorphism ormigmatiza- tion,geothermal bittern,volcanism. The ore- control of deep and giant fault and ductile shear beltand tectono- flash space is emphasized,especially,we should notice the long- term, succession and multistage of the

INDUCTION—POLYMERIZATION OF PREPARING POLYMERS OF REGULAR STRUCTURE

Induction-Polymerization is a novel method for preparing polymers of regular structure. In the presence of a special macromolecular compound, monomer molecules are aggregeted regularly around one or more growth points and polymerized to polymer of regular structure. Regularity of structure of the polyacrylamide prepared by this means has been verified with electron diffractograms, DSC curve and IR spectrum of the polymer.

Modeling of 3D-structure for regular fragments of low similarity unknown structure proteins

Because it is hard to search similar structure for low similarity unknown structure proteins directly from the Protein Data Bank （PDB） database, 3D-structure is modeled in this paper for secondary structure regular fragments （α-Helices, β-Strands） of such proteins by the protein secondary structure prediction software, the Basic Local Alignment Search Tool （BLAST） and the side chain construction software SCWRL3. First, the protein secondary structure prediction software is adopted to extract secondary structure fragments from the unknown structure proteins. Then, regular fragments are regulated by BLAST based on comparative modeling, providing main chain configurations. Finally, SCWRL3 is applied to assemble side chains for regular fragments, so that 3D-structure of regular fragments of low similarity unknown structure protein is obtained. Regular fragments of several neurotoxins are used for test. Simulation results show that the prediction errors are less than 0.06nm for regular fragments less than 10 amino acids, implying the simpleness and effectiveness of the proposed method.

MPS-FEM Coupled Method for Study of Wave-Structure Interaction

Nowadays,an increasing number of ships and marine structures are manufactured and inevitably operated in rough sea.As a result,some phenomena related to the violent fluid-elastic structure interactions(e.g.,hydrodynamic slamming on marine vessels,tsunami impact on onshore structures,and sloshing in liquid containers)have aroused huge challenges to ocean engineering fields.In this paper,the moving particle semi-implicit(MPS)method and finite element method(FEM)coupled method is proposed for use in numerical investigations of the interaction between a regular wave and a horizontal suspended structure.The fluid domain calculated by the MPS method is dispersed into fluid particles,and the structure domain solved by the FEM method is dispersed into beam elements.The generation of the 2D regular wave is firstly conducted,and convergence verification is performed to determine appropriate particle spacing for the simulation.Next,the regular wave interacting with a rigid structure is initially performed and verified through the comparison with the laboratory experiments.By verification,the MPS-FEM coupled method can be applied to fluid-structure interaction(FSI)problems with waves.On this basis,taking the flexibility of structure into consideration,the elastic dynamic response of the structure subjected to the wave slamming is investigated,including the evolutions of the free surface,the variation of the wave impact pressures,the velocity distribution,and the structural deformation response.By comparison with the rigid case,the effects of the structural flexibility on wave-elastic structure interaction can be obtained.

Mineralization and Ore-controlling Implications of Low-angle Faults

Abstract Low-angle faults include those occurring in thrust-nappe structures in a compressive setting and the detachment of metamorphic core complexes in an extensional setting. All low-angle faults have their own particularities. The low-angle fault plays an important role in controlling over some endogenetic metallic ore deposits. Based on studies of the Xiaoban gold deposit, Xinzhou gold deposit, and Longfengchang polymetallic ore deposit, and comparisons with other mines, the authors conclude the ore-controlling implications of low-angle faults as follows. (1) Because of high temperature and high pressure, as well as strong ductile deformation, the internal energy of the elements rises in the large-scale deep ductile low-angle faults, which causes the elements to activate and differentiate from the source rocks, forming ore-bearing hydrothermal solution, and bring mineralization to happen. (2) When rising from depths and flowing along the low-angle faults, the ore-bearing hydrothermal solution will alter and replace the tectonites in the fault zone. The rocks of the hanging side and the heading side differ in lithology, texture and structure, which results in changes or dissimilarities of the physical-chemical conditions. This destroys the balance of the hydrothermal solution system and causes the dissolved ore-forming elements to precipitate; as a result, a deposit is formed. Therefore, the meso-shallow ductile-brittle low-angle faults play the role of a geochemical interface in the process of mineralization. (3) Low-angle faults are often one of the important host structures.

Inverse Load Identification in Stiffened Plate Structure Based on in situ Strain Measurement

For practical engineering structures,it is usually difficult to measure external load distribution in a direct manner,which makes inverse load identification important.Specifically,load identification is a typical inverse problem,for which the models(e.g.,response matrix)are often ill-posed,resulting in degraded accuracy and impaired noise immunity of load identification.This study aims at identifying external loads in a stiffened plate structure,through comparing the effectiveness of different methods for parameter selection in regulation problems,including the Generalized Cross Validation(GCV)method,the Ordinary Cross Validation method and the truncated singular value decomposition method.With demonstrated high accuracy,the GCV method is used to identify concentrated loads in three different directions(e.g.,vertical,lateral and longitudinal)exerted on a stiffened plate.The results show that the GCV method is able to effectively identify multi-source static loads,with relative errors less than 5%.Moreover,under the situation of swept frequency excitation,when the excitation frequency is near the natural frequency of the structure,the GCV method can achieve much higher accuracy compared with direct inversion.At other excitation frequencies,the average recognition error of the GCV method load identification less than 10%.

Analysis on geological structures influencing gas occurrence at Qidong coalmine

The occurrence state of methane is mostly controlled by coalfield geologicalstructures.The coal-bearing strata at Qidong coalmine experienced many tectonic cyclessince their formation.The gas content made by the complicated structural geologic systemat the coalfield is very different, which is obviously higher on the north side of the Weimiaofracture belt than that on the south side and near itself.This thesis discussed the gas occurrenceregularity based on the geometric characteristics of the geological structure andits regional tectonic evolution.This study can provide a foundation for coalfield exploitationand deal with coal and gas outburst.

RGBD Salient Object Detection by Structured Low-Rank Matrix Recovery and Laplacian Constraint

A structured low-rank matrix recovery model for RGBD salient object detection is proposed. Firstly, the problem is described by a low-rank matrix recovery, and the hierarchical structure of RGB image is added to the sparsity term. Secondly, the depth information is fused into the model by a Laplacian regularization term to ensure that the image regions which share similar depth value will be allocated to similar saliency value. Thirdly, a variation of alternating direction method is proposed to solve the proposed model. Finally, both quantitative and qualitative experimental results on NLPR1000 and NJU400 show the advantage of the proposed RGBD salient object detection model. © 2017, Tianjin University and Springer-Verlag Berlin Heidelberg.

Structural Connectivity Enhanced Anisotropic 3D Network for Brain Midline Delineation

Brain midline delineation can facilitate the clinical evaluation of brain midline shift,which has a pivotal role in the diagnosis and prognosis of various brain pathology.However,there are still challenges for brain midline delineation:1)the largely deformed midline is hard to localize if mixed with severe cerebral hemorrhage;2)the predicted midlines of recent methods are not smooth and continuous which violates the structural priority.To overcome these challenges,we propose an anisotropic three dimensional(3D)network with context-aware refinement(A3D-CAR)for brain midline modeling.The proposed network fuses 3D context from different two dimensional(2D)slices through asymmetric context fusion.To exploit the elongated structure of the midline,an anisotropic block is designed to balance the difference between the adjacent pixels in the horizontal and vertical directions.For maintaining the structural priority of a brain midline,we present a novel 3D connectivity regular loss(3D CRL)to penalize the disconnectivity between nearby coordinates.Extensive experiments on the CQ dataset and one in-house dataset show that the proposed method outperforms three state-of-the-art methods on four evaluation metrics without excessive computational burden.

空间结构模型优化的弹性参数FFT-MA随机建模

快速傅里叶变换滑动平均(FFT-MA)法是一种灵活高效的随机建模方法,在地下介质高分辨率建模、复杂介质非平稳建模和不确定性评价等方面具有重要的应用价值。准确构建空间结构模型是利用FFT-MA方法生成合理的随机模型的关键。然而,以往对FFT-MA方法的研究并未提出准确构建空间结构模型的有效方法。为此,文中提出了一种有效的空间结构模型估计方法。该方法基于反演思想,通过最小化随机模型与测井数据和地震数据的空间结构差异,分别估计空间结构模型的纵向自相关长度和横向自相关长度。同时,为了优化空间结构模型的估计效果,在纵向自相关长度的反演过程中引入边界保护正则化,以提高反演的稳定性。此外,将地震约束引入模型优选以提高随机模型的稳定性。实验结果表明:该方法能够稳定估计地下介质的非平稳空间结构模型,从而建立准确描述复杂储层非平稳空间相关特征的高分辨率随机模型。与基于序贯高斯协模拟的随机建模方法相比,使用空间结构模型优化的FFT-MA随机建模方法能够有效刻画多种复杂地质构造从而实现复杂储层建模。

基于因果正则化极限学习机的风电功率短期预测方法

随着风电并网比例的逐年提高,电力系统对风电功率预测的准确性和稳定性提出了更高要求。对于同一风电场而言,为了避免不同特征选择方法所选择的风电场特征子集不同,从因果关系的角度出发,提出了一种基于因果正则化极限学习机(causal regularized extreme learning machine, CRELM)的风电功率短期预测方法。首先将极限学习机(extreme learning machine, ELM)建模为结构因果模型(structural causal model, SCM),在此基础上计算隐藏层神经元与输出层神经元之间的平均因果效应向量。然后将该平均因果效应向量与输出层权重相结合构成因果正则化项,在最小化训练误差的同时最大化网络的因果关系,以进一步提升模型的预测准确性和预测稳定性。最后,以国内蒙西某风电场数据为例,与采用特征选择或不采用特征选择的预测模型相对比,验证了所提方法的有效性和适用性。

The Principle of Square Hole Machining and It＇s Tooling Structure Design

In order to meet the rapid needs of processing square hole in mechanical equipment, the paper expounds the square hole processing method： planetary wheel method, and analyze the principle of tooling structure and process with computer graphics parameters design. The results that, as long as the appropriate parameters, using the above method not only can punch the square hole, can also be processed triangle, the five angle and hexagonal regular polygon holes. The square hole processing method can provide theoretical basis and engineering reliable reference for related engineering and technical personnel.

近距爆炸装药形状对板架结构的损伤规律研究

为了研究近距爆炸时柱形装药长径比对板架结构变形损伤的影响规律,以某典型缩比板架结构作为研究对象,采用经实验验证的有限元模型,分析了5种不同长径比装药爆炸对板架变形损伤的影响,并对比分析了药柱与板架夹角为0°和90°两种情形的区别。研究结果表明:随着装药长径比的增大,夹角为90°情况下冲击波反射超压和冲量逐渐减小,夹角0°情况下冲击波反射超压和冲量逐渐增大,高压区逐渐从装药轴线方向到径向方向移动;板架的残余位移与装药质量近似满足线性关系,装药长径比增大时,夹角为90°时靶板的残余位移逐渐减小,夹角为0°时位移逐渐增大;基于板架中心点最大残余位移D,拟合得到了药柱轴线与靶板表面夹角为90°和0°时的轴向等效形状因子β_(1)和径向等效因子β_(2),当比例距离在0.4~0.7 m/kg^(1/3)范围内时,预测误差小于15%,能较好的预估不同装药长径比下板架的中心点残余位移,获取结构的损伤水平。

面向鲁棒图结构防御的过参数化图神经网络

图数据在现实应用中普遍存在,图神经网络(GNN)被广泛应用于分析图数据,然而GNN的性能会被图结构上的对抗攻击剧烈影响.应对图结构上的对抗攻击,现有的防御方法一般基于图内聚先验进行低秩图结构重构.但是现有的图结构对抗防御方法无法自适应秩真值进行低秩图结构重构,同时低秩图结构与下游任务语义存在错配.为了解决以上问题,基于过参数化的隐式正则效应提出过参数化图神经网络(OPGNN)方法,并形式化证明所提方法可以自适应求解低秩图结构,同时证明节点深层表征上的过参数化残差链接可以有效解决语义错配.在真实数据集上的实验结果表明,OPGNN方法相对于现有基线方法具有更好的鲁棒性,同时,OPGNN方法框架在不同的图神经网络骨干上如GCN、APPNP和GPRGNN上显著有效.

基于双向自举蒸馏的异质云-端医疗对话联邦

医疗对话场景下的数据/模型异质、数据类型不同,为此提出新的联邦学习方法.云模型和端模型以相互自举蒸馏的方式进行知识递进传递.端到云的自举蒸馏过程为多教师-单学生模式,知识被从多个局部模型蒸馏统一到全局模型;云到端的自举蒸馏过程为单教师-多学生模式,知识被从全局模型蒸馏回多个局部模型.在医疗对话ReMeDi和MedDG数据集上,所提方法与经典基线相比通过文本生成指标评价获得了显著提高,训练速度有所提升.

基于RepVGG-A0改进的公路车型识别网络

针对当前车型识别过程中检测精度与实时性难以平衡的问题,提出了一种基于RepVGG-A0改进的公路车型识别网络,利用结构重参数化思想融合多分枝网络以提升网络推理速度。使用混合空洞卷积替换传统卷积,强化了模型对大目标的识别能力。在网络主干中插入融合残差结构的坐标注意力(RES-CA)模块,提升了网络对有效特征信息的提取能力,同时避免了梯度消失与梯度退化造成的影响。此外采用了标签平滑正则化方法对损失函数进行改进,降低了模型过拟合对检测结果的影响,提升了模型的泛化性。经验证,本方法在公路车辆数据集BIT-Vehicle上的识别准确率达到了97.17%,较原模型提升了2.67%,优于现有的ResNet-18,VGG等网络模型,同时保证了模型的检测速度。