钻孔瞬变电磁法扫描探测RCQPSO-LMO组合算法2.5D反演

利用钻孔进行超前探测地质构造及含水体是地下开挖工程中的常规手段,如何利用这些钻孔进行钻孔瞬变电磁法扫描探测,从而实现钻孔孔壁外围地质异常体的精细探测,对实现地下工程地质透明化具有重要的指导意义.本文提出钻孔瞬变电磁法扫描探测2.5D反演的数据解译方法,首先针对随机性反演算法时效性低,易陷入局部最优解,而确定性反演算法依赖初始模型的问题,提出了组合策略的量子粒子群优化算法用来随机搜索最优初始模型.在此基础上,利用Levenberg-Marquarat方法求解Occam反演的目标函数,形成了RCQPSO-LMO组合算法进行2.5D反演,通过对比组合算法和单一算法,验证了组合算法具有更精确的反演结果.其次结合屏蔽条件下扫描探测,对比分析了有无屏蔽的2.5D反演结果,通过设定屏蔽系数对非探测方向信号进行部分压制,可以较好地解决钻孔径向扫描探测中对非探测方向信号部分屏蔽下的反演及成像.最后建立三组理论模型进行组合算法2.5D反演,结果表明：组合算法反演结果与理论模型的一致性较好,对低阻异常体的反演精度较高,验证了组合算法对钻孔孔壁外围低阻异常体具有较高的反演精度和分辨能力.

     

大同地区早中生代煌斑岩—碳酸岩岩墙群

大同地区密集分布的岩墙群由基性－超基性煌斑岩和碳酸岩组成,侵入石炭－二叠纪煤层中。与地层的接触关系和Rb-Sr,K-Ar同位素年代学数据表明墙群的侵位于晚三叠世。Sr,Nd同位素比值显示,碳酸岩和煌斑岩都来源于壳幔混合型岩浆源区。稳定元素比值和相容,不相容元素同等富集则表明是碳酸岩和硅酸盐混合的原始母岩浆液态不混熔的产物。 岩墙群形成的地质条件和构造背景表明,它们是在伸展构造背景下,受到不同级别的构造控制,大同岩墙群的研究对进一步认识华北早中生代岩石圈演化有重要的意义。     

三层堤基中细砂层厚度对管涌影响的试验研究

对由弱透水黏土层、细砂层和强透水砂砾层组成的三层堤基进行了管涌发展的砂槽模型试验,为了便于观察分析,细砂层由各种颜色的细彩砂依次排列在砂砾石层上表面,通过改变彩砂层的厚度分析研究了不同细砂层厚度对管涌发生、发展机制及过程的影响。试验结果表明,三层堤基细砂层厚度的不同使管涌发生的临界水力梯度、涌砂量和通道发展的速度不同,与双层堤基有很大区别。临界水力梯度是由多种元素决定的,包括破坏土体的性质及其整体性等;细砂层的存在使流量在渗透变形初期对涌砂不敏感;在试验中发生的相同水位下多次间歇性涌砂,其原因一方面是颗粒在运动过程中发生堵塞,另一方面是通道边界的土体失去支撑发生应力释放,抵抗力随着时间逐渐减小。     

阿尔泰额尔齐斯带东段酸性岩墙群地球化学特征及其地质意义

野外地质调查发现在阿尔泰南缘额尔齐斯构造带东段、额尔齐斯活动断裂与富蕴-锡泊渡断裂之间发育了一套未变形的酸性岩墙群.岩墙群侵位于海西期片麻岩化花岗岩和上石炭统深变质的额尔齐斯组岩层中,薄片鉴定和岩石地球化学分析确定为具有细晶结构的流纹斑岩.岩石的SiO2含量为70.9%～75.38%,K2O＋Na2O含量为7.55%～8.99%;大部分样品Na2O＞K2O,里特曼指数值为1.8～2.5之间,Al2O3=12.80%～14.53%,A/CNK=0.9～1.1,NK/A=0.7～0.9,具有准铝质-弱过铝质、低镁、高钾钠、低钙和锶、高（Fe）mol/（Mg）mol特点,具A型花岗岩类特点,属于亚碱性脉岩.岩石轻稀土富集,大离子亲石元素相对富集,具有明显的中等负铕异常,壳源特征明显.锆石U-Pb二次粒子微探针测年显示岩墙群形成于277～286Ma.推断岩墙是阿尔泰海西期造山运动结束后,在拉张构造环境下的地壳局部熔融产物.     

Optimal rural land use allocation in central China: Linking the effect of spatiotemporal patterns and policy interventions

Rural land use development is experiencing a transition stage of socioeconomic and land use development in China. Historic land use transition process and policy interventions have key influence on the applicability of land use allocation solutions in future land use management. Strategic land use allocation is therefore required to possess a good adjustment capability to the transition process. Although heuristic optimization methods have been promising to solve land use allocation problems, most of them ignored the spatially explicit effect of historic land use transition and policies. To help resolve this issue, this study aims to optimize future land use pattern in the context of rural land use development. We took Yunmeng County, one of the typical major grain producing and rapidly urbanizing areas in central China, as a case study and solved the sustainable land use allocation problem by using an improved heuristic optimization model. The model was constructed based on the integration of a spatial discrete particle swarm optimization and cellular automata-Markov simulation approach. The spatiotemporal land use patterns and policy interventions were represented by the CA-Markov as in spatially explicit transition rules, and then incorporated into the discrete PSO for optimal land use solutions. We examined the influence of the joint effect of spatiotemporal land use patterns and policy interventions on the land use allocation outcome. Our results demonstrate the robustness and potential of the proposed model, and, more importantly, indicate the significance of incorporating the spatiotemporal land use patterns and policy interventions into rural land use allocation.     

Phreatomagmatic boulder conglomerates at the tip of the ca 2772 Ma Black Range dolerite dyke,Pilbara Craton,Western Australia

Unusual volcanic conglomerates with a mixture of well-rounded granitic boulders (to 1.2 m diameter) derived from adjacent basement rocks, and smaller (1 – 10 cm) subspherical basaltic droplets with chilled margins occupy a linear zone along strike of the northern end of the Late Archaean Black Range dolerite dyke in the Pilbara Craton, Western Australia. The matrix of the volcanic conglomerates becomes more angular with decreasing grainsize and grades to rock flour, a trend opposite to that in sedimentary conglomerates. In other places, the matrix consists of chlorite that cuts through, and resorbs, granitic clasts, indicating an origin as volcanic melt. The volcanic conglomerates have peperitic contacts with immediately adjacent flows of the Mt Roe Basalt of the Fortescue Group. A welded volcanic tuff at the peperitic contact is dated at 2767 ± 3 Ma, within error of the 2772 ± 2 Ma Black Range dolerite dyke and the Mt Roe Basalt (2775 ± 10 Ma), confirming the contemporaneity of formation of these geological elements. Subsequent normal faulting has juxtaposed the higher level conglomerates down into their present exposure level along strike of the Black Range dolerite dyke. The linear zone of volcanic conglomerates is interpreted to represent a phreatomagmatic pebble dyke that formed immediately above, and as a result of intrusion of, the Black Range dolerite dyke. Interaction of magma with groundwater caused phreatomagmatic brecciation of the country rock, in situ milling of granitic boulders, incorporation of basaltic melt droplets, and the formation of a mixed matrix of devitrified volcanic glass and granitic material. This process was accompanied by along-strike epithermal Cu – Hg – Au mineralisation.     

利用人工鱼群算法优化高斯过程模型及应用分析

基于高斯过程利用传统共轭梯度法搜索超参数,存在对初始值依赖性强、获得局部最优解的不足。本文采用人工鱼群算法对超参数进行智能寻优,建立了基于人工鱼群算法的高斯过程模型对变形体形变进行预测分析。通过隧道和基坑两个工程实例计算比对分析,NN、SE和RQ 3种核函数中NN核函数的预测效果最好,平均相对误差分别为0.69%和1.06%。结果表明超参数优化模型的预测精度得到了较大的提高,改善了高斯过程算法本身存在的超参数求解方面的不足,效果明显。     

专题地图色彩的智能设计

为提高专题地图色彩设计的智能化水平,结合专题地图制图的实际需求,给出了一种专题地图色彩智能设计算法;其基本思路为以设计者的配色期望、色彩情感数据库和色彩协调理论为基础,采用粒子群算法进行配色方案的选取,借助孟.斯宾瑟的美度公式对配色方案进行进一步优化,进而得出专题地图智能设色的最优方案。     

基于改进PSO-SVM的噪声稳健建筑物变形监测方法

对高层、超高层建筑物进行实时,高精度的变形监测对提前预防安全隐患,保证人民生命财产安全具有重要意义.建筑物变形作为一种典型的随机性和微弱性过程,噪声等误差的存在会影响从中提取有用的变形信息.针对该问题,提出一种改进粒子群(ParticleSwarmOptimization,PSO)算法优化支持向量机(SupportVe...     

基于粒子群优化神经网络算法的深基坑变形预测方法

深基坑变形预测是进行施工参数调整和确保深基坑施工安全的重要手段,而如何对其变形进行有效、准确的预测是一个有待解决的技术难题。采用粒子群优化算法对神经网络模型的初始权值和阈值进行优化,并将已有的变形监测数据作为神经网络的输入参数,建立了基于粒子群优化神经网络算法的深基坑变形预测方法。将形成的方法应用于长春市火车站北广场深基坑开挖监测工程中。结果表明：8号水平位移测点预测结果的均方根误差为3.78%,平均百分比误差为5.48%;9号地面沉降点预测结果的均方根误差为5.62%,平均百分比误差为3.23%。经验证,本文方法预测深基坑开挖过程中的变形具有较高的可信度。