Prediction of undrained shear strength using extreme gradient boosting and random forest based on Bayesian optimization

Accurate assessment of undrained shear strength(USS)for soft sensitive clays is a great concern in geotechnical engineering practice.This study applies novel data-driven extreme gradient boosting(XGBoost)and random forest(RF)ensemble learning methods for capturing the relationships between the USS and various basic soil parameters.Based on the soil data sets from TC304 database,a general approach is developed to predict the USS of soft clays using the two machine learning methods above,where five feature variables including the preconsolidation stress(PS),vertical effective stress(VES),liquid limit(LL),plastic limit(PL)and natural water content(W)are adopted.To reduce the dependence on the rule of thumb and inefficient brute-force search,the Bayesian optimization method is applied to determine the appropriate model hyper-parameters of both XGBoost and RF.The developed models are comprehensively compared with three comparison machine learning methods and two transformation models with respect to predictive accuracy and robustness under 5-fold cross-validation(CV).It is shown that XGBoost-based and RF-based methods outperform these approaches.Besides,the XGBoostbased model provides feature importance ranks,which makes it a promising tool in the prediction of geotechnical parameters and enhances the interpretability of model.     

An optimization on machine learning algorithms for mapping snow avalanche susceptibility

Natural Hazards - Mapping avalanche-prone areas to mitigate damages is important and vital for safety and development planning. New hybrid models are introduced for snow avalanche susceptibility...     

Nonlinear solver for three-phase transport problems based on approximate trust regions

Implicit transport solvers used in reservoir simulation can take longer time steps than explicit solvers, but for long time steps, the commonly used Newton-Raphson’s method will often fail to converge. The convergence issues may manifest themselves as oscillating residuals even though the implicit discretization itself is stable. This behavior occurs because the fractional flow-type flux functions often change between convex and concave during long time steps, resulting in multiple contraction regions for the Newton-Raphson solver. The common strategy to overcome this is to set limits on the saturation changes during the nonlinear iteration, but such a limit has to be determined on a case by case basis, excess iterations may be required, and practical convergence is not guaranteed for a given problem. Previous work on this problem by multiple authors has resulted in solvers based on trust regions, where unconditional convergence can be obtained for incompressible two-phase flow provided a priori analytical knowledge of the flux function exists. The goal of our work is to extend this methodology to a solver where inflection points demarking the different contraction regions do not need to be explicitly known. Instead, these values are estimated during the solution process, giving improved convergence by a local computation for each interface in the simulation model. By systematically reducing updates over regions known to produce convergence issues, it is possible to greatly reduce the computational expense, making the same formulation suitable for an arbitrary number of components. We present a series of numerical results, including arbitrary time-step lengths for two and three-phase gravity segregation, as well as three-dimensional gas and water injection problems with wells and a mixture of both viscous and gravity-dominated flow regimes. The test cases are a systematic validation on a wide variety of both analytical and tabulated relative permeability curves.     

基于遗传算法的巷道位移反分析研究

将遗传算法应用到位移反分析研究中,采用遗传算法与FLAC数值软件相结合的手段,研制了弹塑性巷道位移反分析程序。对该程序测试的结果表明,用遗传算法这一模拟生物进化的方法,进行巷道位移反分析,可以同时反演E,μ,C,ф,σx,σy,τxy7个参数,反演精度较高。     

基于连续不确定区域的蚁群算法求解水文地质逆问题

本文在介绍求解复杂组合优化问题而提出的蚁群算法的基础上,提出一种基于连续不确定区域的蚁群算法。该方法将连续变量优化问题转化为一个组合优化问题,进而将蚁群算法求解。通过两个水文地质逆问题的求解,表明该方法在参数识别中计算结果可靠,具有较好的计算效率。     

基于非结构化网格的重力梯度张量反演

由于重力梯度张量是重力位的二阶导数,与传统的重力测量相比,重力梯度张量具有更高的分辨率,并且重力梯度张量具有5个相互独立的分量,包含了更多的地下空间信息和密度信息,因此将重力梯度张量数据应用到反演中可以得到较好的反演结果。非结构化网格具有很强的几何适应性,能够较好地拟合复杂异常体的边界,通过非结构化网格对反演目标区域进行离散,可以降低剖分误差,从而提高计算精度。为了降低反演过程中的多解性问题,将地球物理反问题的广义正则化目标函数应用于基于非结构化网格的三维重力梯度张量反演中,推导了相关公式,实现了各分量的独立反演,并阐述了深度加权函数在反演过程中的作用。为了充分利用重力梯度张量各分量所携带的密度、空间信息,将5个独立的分量进行了联合反演,反演结果表明,基于非结构化网格的三维重力梯度张量反演能够较好地反映地下异常源的物性分布和赋存位置。通过与长方体网格反演结果对比,本反演方法突出了非结构化网格反演的优越性;最后,通过较复杂组合模型的计算证实了方法的实用性。     

基于共轭梯度法的全波形反演

全波形反演是一种全新的地震成像方法,主要利用全波形信息反演地下介质参数,通过非线性优化波场理论值和观测值的残差实现波形反演。基于时间域声波方程,建立了时间域波场残差目标函数,以分层模型为例,分别从波场精度、目标函数收敛性和运行时间3个方面,比较了共轭梯度（Conjugate Gradient,CG）算法和拟牛顿算法（Broyden-Fletcher-Goldfarb-Shanno,BFGS）反演的效果。同时,应用共轭梯度法对正、逆断层模型和Marmousi模型进行了速度结构反演。反演结果表明：共轭梯度法计算效率较高,反演得到的速度模型精度更高,反演效果较好,是一种有效的波形反演方法。     

基于导数迭代法的重力归一化总梯度

针对传统重力归一化总梯度方法对油气藏分辨率不够和深度计算精度较低的问题,本文利用导数迭代法的向下延拓结果,计算重力归一化总梯度。将传统方法和本文方法同时用于模型试验,分析谐波数N、埋深和充满系数对G~H场特征的影响。结果表明,相比于传统方法,本文方法受谐波数N的影响较弱,G~H场对油气藏的分辨率更高,反演的油气藏深度更加准确。通过实际应用,证实了本文方法的有效性。     

基于分形与损伤理论的岩石声-应力相关性理论模型研究

建立单一微裂纹在单轴压缩荷载作用下的力学模型,采用宏观复合型断裂理论中的最大周向应力理论,求出单一微裂纹达到稳定时的形态尺寸.利用分形损伤理论,求出不同荷载阶段的分形维数;由分形维数的定义导出了不同荷载阶段的裂纹个数的变化情况;结合单一微裂纹的形态尺寸,求出不同荷载阶段中岩石孔隙率与荷载值之间的关系;利用目前被广泛采用的超声波纵波速度与孔隙率的关系式,最终导出单轴压缩荷载作用下岩石超声波纵波速度与应力的理论关系式.进行岩石声-应力的试验测试,根据试验成果作出声-应力曲线,并与理论声-应力曲线进行比对,结果表明,理论模型比较适合于反映岩石受载全过程中的后半部分的声-应力关系.     

Limit equilibrium method based on an approximate lower bound method with a variable factor of safety that can consider residual strength

In this paper, the factor of safety for a prescribed slip surface will be determined from an equivalent lower bound method (extremum principle), which can satisfy all equilibrium conditions without an interslice force function. This approach will give an overall factor of safety close to that from the classical methods for normal problems, while the thrust line, the local factor of safety for individual slice/block and the progressive yielding phenomenon can be estimated, which will be useful for some special cases. The force and moment equilibrium of every slice will be satisfied, while the location of the thrust line will always be acceptable in the present formulation. To solve the difficult optimisation problem, an innovative coupled particle swarm/harmony search algorithm is proposed in this paper, and a practical engineering problem for which the factor of safety is close to 1.0 is used to illustrate the consideration of the residual strength in the limit equilibrium slope stability analysis.     

Mixed multiscale finite element methods using approximate global information based on partial upscaling

The use of limited global information in multiscale simulations is needed when there is no scale separation. Previous approaches entail fine-scale simulations in the computation of the global information. The computation of the global information is expensive. In this paper, we propose the use of approximate global information based on partial upscaling. A requirement for partial homogenization is to capture long-range (non-local) effects present in the fine-scale solution, while homogenizing some of the smallest scales. The local information at these smallest scales is captured in the computation of basis functions. Thus, the proposed approach allows us to avoid the computations at the scales that can be homogenized. This results in coarser problems for the computation of global fields. We analyze the convergence of the proposed method. Mathematical formalism is introduced, which allows estimating the errors due to small scales that are homogenized. The proposed method is applied to simulate two-phase flows in heterogeneous porous media. Numerical results are presented for various permeability fields, including those generated using two-point correlation functions and channelized permeability fields from the SPE Comparative Project (Christie and Blunt, SPE Reserv Evalu Eng 4:308–317, 2001). We consider simple cases where one can identify the scales that can be homogenized. For more general cases, we suggest the use of upscaling on the coarse grid with the size smaller than the target coarse grid where multiscale basis functions are constructed. This intermediate coarse grid renders a partially upscaled solution that contains essential non-local information. Numerical examples demonstrate that the use of approximate global information provides better accuracy than purely local multiscale methods.     

亚塑性模型不同积分算法的数值实现

亚塑性模型是以Jaumann应力率张量及变形率张量描述的一种率型本构关系,本构关系在非线性有限元分析计算中具有关键作用,解决用应变增量求解应力响应的问题需要一个时间积分过程。针对亚塑性本构模型发展了自适应隐式和显式两种不同的积分算法,给出了误差控制的方法,同时推导了自适应隐式积分算法所需的一致切向模量,并采用了两个不同的单元,利用ABAQUS平台比较了两种积分算法的数值模拟结果。为了实现从ABAQUS/Standard到ABAQUS/Explicit的过渡,开发了UMAT-VUMAT接口,从而可以使已有的UMAT子程序用于大变形动力问题分析。算例分析证明了研究结果的正确性。     

基于深度学习的智能矿物识别方法研究

矿物识别在许多研究领域都有着重要作用,基于深度学习技术的智能矿物识别为这些领域带来了新的发展方向,不仅能有效节省人工成本,还能减小识别错误。针对石英、角闪石、黑云母、石榴石和橄榄石共5种矿物进行实验,提出了一种准确高效的智能矿物识别方法。实验采用图像分析常用的卷积神经网络建立模型,设计出一套基于残差神经网络的矿物识别方法。本实验独立采集了5种矿物的偏光显微图像数据集,用于模型的训练、验证和测试,并通过合理的数据增强策略来扩充训练数据集。在卷积神经网络的结构设计上,选取了ResNet-18作为框架,最终于模型测试中取得89%的准确率,成功训练出一个较为精准的矿物识别模型,实现了基于深度学习的智能矿物识别方法。     

Evolutionary connections of biological kingdoms based on protein and nucleic acid sequence evidence

Evolutionary trees are derived based on nucleic acid and protein sequences and the methods of numerical taxonomy. From 5S ribosomal RNA sequences, molecules that participate in protein synthesis, the order of branching of many prokaryote species and the nuclear and organellar components of eukaryote species are inferred. All oxygen-evolving forms, including the chloroplasts and blue-green algae, are to be found on a single branch; the mitochondria are found with the Rhodospirillaceae, nonsulfur photosynthetic bacteria; and the eukaryote cytoplasmic components are found with the archae-bacteria, including Thermoplasma and Sulfolobus. This tree clearly supports the symbiotic origin of the eukaryote organelles. There is much redundant information available from c-type cytochromes, cytochromes c′, 4Fe-4S ferredoxins and 5.8S rRNA molecules. A gene duplication of ferredoxin, probably predating all of the species divergences, permitted us to infer an ancestral molecule that can be placed on the tree to give the base. The heterotrophic anaerobes have a sequence most resembling the ancestral sequence, suggesting that the ancestor may have been a heterotroph. Ten prokaryote species occur on both the c-type and the c′-type cytochrome trees. There is only one inconsistency in the topologies of the two trees, which may reflect a gene transfer among prokaryotes confirming that, for these genes of fundamental importance, gene transfer followed by acceptance is rare. The eukaryote trees derived from 5S and 5.8S rRNA suggest that the earliest eukaryotes were unicellular and that the multicellular forms of plants, animals, fungi and slime molds arose independently.     

Study of threshold gradient for compacted clays based on effective aperture

Compacted clay liner plays a crucial role in preventing leakage and migration of pollutants. Clay often appears in Non-Darcy phenomenon, one of which is characterized by having the threshold gradient. The threshold gradients exhibited by different clays are usually quite different. The reason for the discrepancy is currently unknown. To solve this problem, the mineral compositions of two clays were selected for test. The tests included threshold gradient, aperture and the pore water content. The test results showed that kaolin-based Clay A had no threshold gradient, whereas illite and Montmorillonite-based Clay B presented the threshold gradient that decreased with the increase of the porosity. For the samples mixed with Clay A and Clay B, the values of the threshold gradient increased with increasing the content of Clay B in the mixed samples. The reason is that the porosity and mineral content can affect the bound water content. If the value of pF is greater than 3.8, the bound water is gelatinous “immobile water”. Based on the water retention curves, the effective porosity can be obtained by subtracting the volume occupied by “immobile water” from the void of the soil particles. Then the modified effective aperture can be obtained by Hangen-Poiseuille law. The modified effective aperture will decrease with the decrease in the porosity and the increase of the content of illite and montmorillonite. This is the key reason why the different clays have different values of the threshold gradients.     

基于BOTDR的光纤应变与顶板沉降变形关系的模型构建与试验研究

可以实现分布式、长距离、实时监测的BOTDR监测技术已在土木工程领域得到了较为广泛的应用,但在岩土工程领域的应用研究起步较晚,相关技术还有待于进一步完善。以布里渊BOTDR的巷道/隧道应变监测技术为基础,研究了沿巷道轴向进行一维布设的光纤应变测量值与隧道顶板沉降变形之间的关系。首先,根据巷道/隧道顶板的变形规律,建立了光纤沿巷道/隧道轴向发生的应变与巷道/隧道顶板的沉降变形之间的数学模型,即圆弧模型、抛物线模型和三角形模型。其次,对巷道/隧道顶板的沉降变形与光纤应变之间的关系进行了试验研究。对比分析表明,理论计算值与试验值具有很好的一致性。利用圆弧模型、抛物线模型和三角形模型等3个计算模型,可以将光纤应变与顶板沉降联系起来,这对于BOTDR技术的应用与推广具有重要的理论意义和工程实用价值。     

Parameters optimization on DHSVM model based on a genetic algorithm

Due to the multiplicity of factors including weather, the underlying surface and human activities, the complexity of parameter optimization for a distributed hydrological model of a watershed land surface goes far beyond the capability of traditional optimization methods. The genetic algorithm is a new attempt to find a solution to this problem. A genetic algorithm design on the Distributed-Hydrology-Soil-Vegetation model (DHSVM) parameter optimization is illustrated in this paper by defining the encoding method, designing the fitness value function, devising the genetic operators, selecting the arithmetic parameters and identifying the arithmetic termination conditions. Finally, a case study of the optimization method is implemented on the Lushi Watershed of the Yellow River Basin and achieves satisfactory results of parameter estimation. The result shows that the genetic algorithm is feasible in optimizing parameters of the DHSVM model.     

基于栅格DEM的多流向算法述评

基于栅格DEM的流向算法是数字地形分析的重要研究内容之一。在分布式水文模型、土壤侵蚀等领域中,当需要获取与流向直接相关的水文参数(如汇流面积、地形指数等)的详细空间分布时,多流向算法(MFD)明显优于单流向算法(SFD)。根据所采用的水流分配策略,将现有MFD分为四类:(1)固定水流分配权重的MFD;(2)水流分配权重随汇流面积变化的MFD;(3)水流分配权重随局域地形特征变化的MFD;(4)基于局域形态单元的MFD。从模型合理性、算法复杂度和易用性、对DEM误差的敏感性等方面进行的分析表明,水流分配权重随局域地形特征变化的MFD优于其他三类MFD。目前对MFD采用人造DEM数据集的直接定量评价研究还十分不足,文章以一个例子来体现直接定量评价MFD的可行性及有效性,同时也定量反映出水流分配权重随局域地形特征变化的MFD具有更低的计算误差。目前通用的针对流向算法的DEM洼地平区预处理算法多针对SFD而设计,不适合MFD的实际应用,本文对适用于MFD的DEM预处理算法所应具有的特征也进行了讨论。     

水流分配策略随下坡坡度变化的多流向算法

提出了一种新的多流向算法方法(MFD-fg),其基本思想是水流分配策略应在空间上随着与下坡坡度相关的地形参数而变,从而地形对水流分配的影响可以得到合理、有效的建模。比较分析表明,MFD-fg采用以最大下坡坡度的线性函数对水流分配进行加权,是较好的水流分配策略。应用实验表明,与目前具代表性的单流向算法与多流向算法相比,MFD-fg方法的结果可获得更合理的汇流面积。