基于Elman型神经网络的金川二矿地表岩移时序预测 模型

建立于煤矿开采基础之上的矿山开采沉陷理论和预测方法并不适用于象金川这样厚大、陡倾的金属矿床开采的岩移问题,因此,本文探讨利用神经网络来对地表岩移进行预测。根据Elman神经网络能够逼近任意非线性函数的特点和具有反映系统动态特性的能力,提出了利用Elman神经网络建立地表岩移时序预报模型的方法。利用金川二矿区GPS监测所得到的时间序列数据,通过对Elman神经网络模型预测值与GPS实测值之间的比较,结果表明模型预测显示了良好的准确性,特别是在时间步长较短情况下,应用于实际预测一定程度上可以弥补金属矿山岩移预测方法不足的缺憾。     

Application of HYDRUS-1D in understanding soil water movement at two typical sites in the North China Plain

Recharge and discharge, such as rainfall infiltration and evapotranspiration in vertical direction, are major processes of water cycle in the shallow groundwater area of the North China Plain. During these processes, soil water movement in the unsaturated zone plays an important role in the transformation from rainfall infiltration to groundwater. The soil water movement models were developed by using HYDRUS-1D software at two typical experimental sites in Cangzhou (CZ) and Hengshui (HS) with different soil, vegetation and similar climate conditions. As shown in the results, the comparison in precipitation infiltration features between the two sites is distinct. The soil water experiences strong evaporation after precipitation infiltration, which accounts for 63% of the total infiltration at the HS site where the soil is homogenous. It is this strong evaporation effect that leads to slight increase of soil water storage. At the CZ site, where the soil is heterogeneous, the evaporation effect exists from July to October of the simulation period. The total evaporation accounts for 33% of the total infiltration, and the evaporation rate is slow. At the end of the simulation period, the soil water storage increases and the water table decreases, indicating a strong storage capacity at this site.     

基于流动电位正演模型的非饱和带水流过程监测

为了获取非饱和带水流过程的信息,借助流动电位正演模型,通过数值实验探讨非降雨和降雨两种条件下非饱和带流动电位和水流过程的关系,然后用南京中山植物园试验场地野外观测的流动电位和张力数据加以对比和验证。野外试验表明:流动电位可以有效地反映非饱和带水流过程。在夏季无降雨入渗的条件下,日周期变化的地表地下温度差导致水分的运动,流动电位准确地指示了非饱和带含水量和毛细压力的变化情况,从而指示出了水分运移的方向;在夏季有降雨入渗的条件下,降雨锋面推进之处,含水量和流动电位同时有明显的响应,进而根据不同位置的流动电位对降雨入渗响应的时刻差,直接求出入渗锋面的推进速度。     

Conversion relationship of rainfall-soil moisture-groundwater in Quaternary thick cohesive soil in Jianghan Plain,Hubei Province,China

The scientific field test site of rainfall-soil moisture-groundwater conversion in Dabie Mountain Area–Jianghan Plain is located in the northern region of the Jianghan Plain, the transition zone between the Dabie Mountain Area and Jianghan Plain. It’s a great field test site to study the material and energy exchange among rainfall, soil moisture, and groundwater of the Earth ’s critical zone in subtropical monsoon climate plain areas. This paper analyzed the connection between rainfall and volume water content (VWC) of soil at different depths of several soil profiles, and the dynamic feature of groundwater was discussed, which reveals the rainfall infiltration recharge of Quaternary Upper Pleistocene strata. The results show that the Quaternary Upper Pleistocene aquifer groundwater accepts a little direct rainfall recharge, while the lateral recharge is the main supplement source. There were 75 effective rainfall events among 120 rainfall events during the monitoring period, with an accumulated amount of 672.9 mm, and the percentages of effective rainfall amount and duration time were 62.50% and 91.56%, respectively. The max evaporation depth at the upper part in Quaternary cohesive soil was no less than 1.4 m. The soil profile was divided into four zones: (1) The sensitive zone of rainfall infiltration within 1.4 m, where the material and energy exchange frequently near the interface between atmosphere and soil; (2) the buffer zone of rainfall infiltration between 1.4 m and 3.5 m; (3) the migration zone of rainfall infiltration between 3.5 m and 5.0 m; and (4) the rainfall infiltration and groundwater level co-influenced zone below 5.0 m. The results revealed the reaction of soil moisture and groundwater to rainfall in the area covered by cohesive soil under humid climate in Earth ’s critical zone, which is of great theoretical and practical significance for groundwater resources evaluation and development, groundwater environmental protection, ecological environmental improvement, drought disaster prevention, and flood disaster prevention in subtropical monsoon climate plain areas.     

Analysis of time-varying rainfall infiltration induced landslide

A case study of rainstorm-induced landslide is modeled following the hourly rainfall time series from the stage of infiltration caused by induced slope movement and soil saturated to excess pore pressures—Transient Rainfall Infiltration and Grid-based Regional Slope-Stability Analysis (TRIGRS). The grid-based landslide stability analysis was conducted to model the increased pore pressures and runoff in the study area under the specified rainfall conditions. The generated time variances of pore pressures help determine landslide characteristics and mechanisms under rainfall conditions. Inputs of soil properties and permeability parameters for landslide stability analysis in the study area were prepared by TRIGRS adopted for transient infiltration analysis. Results of the analyses show that under heavy rainfall conditions, the infiltrated slope is unstable and the time of debris masses movement initiated is correlated to the recorded time. In the initiated landslide, characteristics and effects are considered and reflected in the numerical modeling under combinations of topography, land use, climatic and geological conditions. Results reveal that there is a plane failure surface and a potential circular failure surface at the study site besides the rock topple failures in the crest. A grid-based slope-stability analysis incorporated with the GIS spatial functions is more advantageous than the traditional two-dimensional analysis for specified slope profiles to determine the whole behavior of a slope.     

Response of hydrothermal activity in different types of soil at ground surface to rainfall in permafrost region

In order to understand the hydrothermal activity mechanism of active layers to rainfall in permafrost regions caused by humidification of climate, the differences of ground surface energy balance and hydrothermal activity in different types of shallow soil with the consideration of rainfall were discussed. Based on the meteorological data in 2013 observed at Beiluhe observation station of Tibet Plateau, three types of shallow ground soil (i.e., sandy soil, sandy loam and silty clay) were selected to compare the differences in the water content and energy balance at the ground surface, dynamic processes of water and energy transport in active layers and coupling mechanism under rainfall condition in the plateau using a coupled water-vapor-heat transport model. The results show that the increase of soil particle size leads to the increase of surface net radiation and latent heat of evaporation, but the decrease of soil heat flux. The difference of surface energy balance, especially the sensible heat flux and latent heat of evaporation, are larger in the warm season but smaller in the cold season. The liquid water transport under hydraulic gradient and the water-vapor transport under thermal gradient are obvious as the particle size in soil increases. However, the water-vapor flux under thermal gradient increases but the liquid water flux under hydraulic potential gradient decreases. As a result, the water content in shallow soil decreases accordingly but it increases slightly at the depth of 25 ~75 cm. Moreover, with the increase of soil particle size, the thermal conductivity of soil, convective heat transfer under rainfall and surface evaporation increase, but the soil heat conduction flux and soil temperature gradient decrease. Thus, soil temperature in sandy soil is much higher than that of sandy loam and silty clay at the same depth. The permafrost table declines with the increase of the thickness of active layer, which is unfavourable to permafrost stability. The results can provide theoretical reference for stability prediction and protection of permafrost caused by humidification of climate.     

渭河流域西部季节冻融对浅层非饱和土壤水热变化的影响

以黄土高原渭河流域西部黄土丘陵沟壑区为研究区域,建立了野外观测场地,对该区域浅层非饱和土体冻融过程及水热运移规律对气候作用的响应过程进行了研究与分析。结果表明：气温对地温及地温变幅的影响随深度增加而迅速衰减,地温振幅随深度增加按指数规律衰减且温度波的相位随深度的增加而滞后,地表下200 cm深度以内地温振幅受气温影响较大。该区域裸露地表土壤的最大冻结深度在20～50 cm之间。在土壤冻结过程中,深层土壤未冻水逐渐向冻结层运移,导致深层含水量逐渐减少。不同深度土壤冻结系数随土壤深度的增加而减小,融化系数则相反。地表下50 cm深度以内的土体含水量受降水影响波动显著。土壤含水量与温度呈相似变化,地温峰值出现的时间总滞后于土壤水分,其变异程度均随土壤深度的增加而减小。     

降雨对青藏高原多年冻土区地表辐射的影响——以北麓河地区为例

青藏高原地表辐射变化影响多年冻土地表能量收支平衡、热工计算温度场边界条件以及冻土热稳定性。降雨变化是影响地表辐射的重要因素,以青藏高原北麓河地区的地表辐射资料和浅层地表水热监测数据为基础,分析了降雨对多年冻土区地表辐射特征的影响。研究表明：受降雨影响,辐射分量具有明显的日变化和年际变化特征,短波辐射被严重削弱,地面长波辐射削弱程度较低,而太阳长波辐射增强。小雨作用后,太阳短波、地面短波、地面长波辐射日积分量分别减少24.6%、37.9%、4.2%,太阳长波增加了4.3%;中雨作用后,太阳短波、地面短波、地面长波辐射日积分量分别减少32.2%、43.4%、1.7%,太阳长波增加了11.6%;大雨作用后,太阳短波、地面短波、地面长波辐射日积分量分别减少56.3%、65.5%、4.4%,太阳长波增加了10.7%;降雨作用后地表反照率减小,且呈不对称“U”形变化,地表净辐射增加,长期时间尺度上降雨对净辐射影响不明显;降雨对辐射特征的影响程度与降雨强度呈正相关且地表辐射对降雨变化的响应有滞后效应。在暖湿化背景下,冻土热工计算中应考虑降雨对多年冻土区地表能量平衡和水热边界的影响。     

脉冲地震动中竖向加速度对场地液化的影响

为研究近断层脉冲地震动中竖向加速度对砂土场地液化的影响,基于有限元平台OpenSees开发的边界面塑性本构模型,建立了动单剪单元试验模型和饱和砂土三维有限元模型。选取台湾Chi-Chi地震中10条具有速度脉冲特性的地震波,对比分析了水平双向脉冲波与三向脉冲波作用下土柱竖向位移、循环应力比、孔压比及等效循环周数的差异性,继而明确了脉冲地震动中竖向加速度对砂土液化的影响规律。研究表明,三向脉冲地震波中竖向加速度分量对场地永久位移值影响较小,但使永久位移的发展持时明显增大;土柱循环应力比受竖向地震动影响较小,因此分析脉冲地震动对场地剪切特性的影响时,可将三向脉冲地震动简化为水平双向地震动;考虑竖向地震动的三向脉冲地震波引起的孔压比变化幅度较大,孔压消散时间较长;三向脉冲地震波对应的等效循环周数较大,地震动发展持时长,可认为竖向加速度对场地液化有促进作用。     

Development of an inverse analysis framework for extracting dynamic soil behavior and pore pressure response from downhole array measurements

Observations from earthquakes over the past several decades have highlighted the importance of local site conditions on propagated ground motions. Downhole arrays are deployed to measure motions at the ground surface and within the soil profile, and also to record the pore pressure response within the soft soil profiles during excitation. The degradation of soil stiffness as excess pore pressures are generated during earthquake events has also been observed. An inverse analysis framework is developed and demonstrated to directly extract soil material behavior including pore water pressure (PWP) generation from downhole array measurements that can then be readily used in 1D nonlinear site response analysis. The self‐learning simulations (SelfSim) inverse analysis framework, previously developed for total stress site response analysis, is extended to extract PWP generation behavior of the soil in addition to cyclic response during ground shaking. A Neural Network based constitutive model is introduced to represent PWP generation during cyclic loading. A new analysis scheme is introduced that can use data from co‐located piezometer and accelerometer sensors. The successful performance of the proposed framework is demonstrated using four synthetic vertical array recordings. Copyright © 2013 John Wiley & Sons, Ltd.     

Evolution of soil-water states in the vadose zone of a desert soil after an extreme rainfall event and its impact on the ecosystem

Evolution of soil-water movement patterns following rare and extreme rainfall events in arid climates is not well understood, but it has significant effects on water availability for desert plants and on the hydrological cycle at small scale. Here, field data and the Hydrus-1D model were used to simulate the mechanisms of soil water and vapor transport, and the control factors associated with temporal variability in the soil water and temperature were analyzed. The results showed that thermal vapor transport with a no rainfall scenario determined daily variability in water content at the soil surface. During rainfall, isothermal liquid water fluctuated as a result of dry sandy soils and matric potential in the upper soil (0–25 cm), and thermally driven vapor played a key role in soil-water transport at 40–60 cm soil depth. After an extreme rainfall event, thermal vapor flux increased and accounted for 11.8% of total liquid and vapor fluxes in daytime with a steep temperature gradient; this was very effective in improving long-term soil-water content after the rain. The simulated results revealed that thermal water vapor greatly contributed to the soil-water balance in the vadose zone of desert soil. This study provided an alternative approach to describing soil-water movement processes in arid environments, and it increased understanding of the availability of water for a desert plant community.

     

Load-settlement behavior modeling of single piles using artificial neural networks and CPT data

Pile foundations are usually used when the conditions of the upper soil layers are weak and unable to support the super-structural loads. Piles carry these super-structural loads deep into the ground. Therefore, the safety and stability of pile-supported structures depends largely on the behavior of the piles. In addition, accurate prediction of pile behavior is necessary to ensure appropriate structural and serviceability performance. In this paper, an ANN model is developed for predicting pile behavior based on the results of cone penetration test (CPT) data. Approximately 500 data sets, obtained from the published literature, are used to develop the ANN model. The paper compares the predictions obtained by the ANN with those given by a number of traditional methods and it is observed that the ANN model significantly outperforms the traditional methods. An important advantage of the ANN model is that the complete load-settlement relationship is captured. Finally, the paper proposes a series of charts for predicting pile behavior that will be useful for pile design.     

地下商场结构对地面运动影响的振动台试验研究

为了研究地下空间结构对场地地震反应的影响，以一个三层地下商场结构为原型，设计并开展了土?地下空间结构动力相互作用振动台模型试验，研究了在6条不同卓越频率输入地震波在不同峰值加速度下场地地震反应的变化规律。试验研究结果表明：（1）地表水平加速度放大系数以结构中轴线为对称轴呈对称分布，地下结构显著影响上方及邻近场地的地表水平加速度。同时，由于地下空间结构的局部场地效应影响，即使在水平地震输入下也将导致场地产生竖向振动，地表竖直加速度放大系数呈M形分布。地下空间结构的影响范围可达到其两侧各一倍宽度距离。（2）地表加速度放大系数受输入地震波卓越频率的影响明显，且随着输入地震波的峰值加速度（PGA）的增大而降低，但空间放大系数随着PGA的增大而增大。（3）地表加速度Fourier谱的峰值频率受输入地震波的卓越频率、场地水平方向的固有频率和竖向固有频率的综合影响。     

Risk degree of debris flow applying neural networks

A number of methods for prediction of debris flows have been studied. However, the successful prediction ratios of debris flows cannot always maintain a stable and reliable level. The objective of this study is to present a stable and reliable analytical model for risk degree predictions of debris flows. This study proposes an Artificial Neural Networks (ANN) model that was constructed by seven significant factors using back-propagation (BP) algorithm. These seven factors include (1) length of creek, (2) average slope, (3) effective watershed area, (4) shape coefficient, (5) median size of soil grain, (6) effective cumulative rainfall, and (7) effective rainfall intensity. A total of 171 potential cases of debris flows collected in eastern Taiwan were fed into the ANN model for training and testing. The average ratio of successful prediction reaching 99.12% demonstrates that the presented ANN model with seven significant factors can provide a highly stable and reliable result for the prediction of debris flows in hazard mitigation and guarding systems.     

Three‐dimensional responses of buried corrugated pipes and ANN‐based method for predicting pipe deflections

This study investigated localized responses, such as circumferential stresses, on corrugation and pipe deflections. Also, this study examined the effect of corrugation geometry on the overall and localized response of corrugated pipes with refined three‐dimensional modeling of the entire soil–pipe interaction system, including corrugation. To investigate the availability of the traditional two‐dimensional method, the results from the three‐dimensional finite element method (FEM) were compared with those from the two‐dimensional FEM. The soil–pipe modeling techniques of this study were verified by comparing the FEM results by Utah State University and analytical results. An artificial neural network (ANN)‐based model to predict vertical deflections of buried corrugated pipes was developed to overcome the shortcomings of existing methods and obtain results that are close to the level of accuracy of FEM results. In order to train an ANN, analyses on a large amount of data were executed with various standardized pipe geometries and burial depths regulated by the Korea Highway Corporation using the two‐dimensional FEM verified in this study. The widely used back propagation algorithm was adopted. The ANN‐based model developed in this study was shown to be an effective tool by comparing the results with test data and sensitivity analyses were executed based on the data from the developed ANN. Copyright © 2010 John Wiley & Sons, Ltd.     

A neural network approach for attenuation relationships: An application using strong ground motion data from Turkey

This paper presents an application of neural network approach for the prediction of peak ground acceleration (PGA) using the strong motion data from Turkey, as a soft computing technique to remove uncertainties in attenuation equations. A training algorithm based on the Fletcher–Reeves conjugate gradient back-propagation was developed and employed for three sample sets of strong ground motion. The input variables in the constructed artificial neural network (ANN) model were the magnitude, the source-to-site distance and the site conditions, and the output was the PGA. The generalization capability of ANN algorithms was tested with the same training data. To demonstrate the authenticity of this approach, the network predictions were compared with the ones from regressions for the corresponding attenuation equations. The results indicated that the fitting between the predicted PGA values by the networks and the observed ones yielded high correlation coefficients (R²). In addition, comparisons of the correlations by the ANN and the regression method showed that the ANN approach performed better than the regression. Even though the developed ANN models suffered from optimal configuration about the generalization capability, they can be conservatively used to well understand the influence of input parameters for the PGA predictions.     

广义回归神经网络预测加筋土支挡结构高度

土工合成材料加筋支挡结构（Geosythetics-Reinforced Retaining Wall, 简称GRW）设计方法主要是建立在似粘聚力理论基础之上的半经验设计法。由于土性及加筋机理的复杂性，常常要对它们进行人为假定，导致计算结果差强人意。神经网络方法与传统方法的不同之处在于不需要主观假定，而是模拟人脑思维，通过数据样本的学习来获得预测结果。引入神经网络技术来预测加筋土支挡结构的设计高度是一种新尝试。由于本问题具有样本容量非常有限、影响因素复杂多样的特点。因此，采用适用于稀土样本数据的广义回归网络（General Regression Neural Network)来预测加筋土支挡结构设计高度。基于MATLAB神经网络工具箱及文献[1]的挡墙离心模型试验结果，建立了一个可用于加筋支挡结构设计高度预测的GRNN网络。通过对足尺试验，实际工程及模型试验结果的检验，表明网络的学习是成功的，具有一定指导意义。     

Equivalent-Linear Seismic Ground Response Analysis of Some Typical Sites in Mumbai

Any earthquake event is associated with a rupture mechanism at the source, propagation of seismic waves through underlying rock and finally these waves travel through the soil layers to the particular site of interest. The bedrock motion is significantly modified at the ground surface due to the presence of local soil layers above the bedrock beneath the site of interest. The estimation of the amplifications in ground response due to the local soil sites is a complex problem to the designers and the problem is more important for mega cities like Mumbai in India, where huge population may get affected due to devastations of earthquake. In the present study, the effect of local soil sites in modifying ground response is studied by performing one dimensional equivalent-linear ground response analysis for some of the typical Mumbai soil sites. Field borelog data of some typical sites in Mumbai city viz. Mangalwadi site, Walkeswar site, BJ Marg near Pandhari Chawl site are considered in this study. The ground responses are observed for range of input motions and the results are presented in terms of surface acceleration time history, ratio of shear stress to vertical effective stress versus time, acceleration response spectrum, Fourier amplitude ratio versus frequency etc. The typical amplifications of ground accelerations considering four strong ground motions with wide variation of low to high MHA, frequency contents and durations are obtained. Results show that MHA, bracketed duration, frequency content have significant effects on the amplification of seismic accelerations for typical 2001 Bhuj motion. The peak ground acceleration amplification factors are found to be about 2.50 for Mangalwadi site, 2.60 for Walkeswar site and 3.45 for BJ Marg site using 2001 Bhuj input motion. The response spectrum along various soil layers are obtained which will be useful for designers for earthquake resistant design of geotechnical structures in Mumbai for similar sites in the absence of site specific data.     

现场试坑浸水试验在湿陷性黄土地区地铁工程中的应用

大厚度湿陷性黄土地基处理是黄土地区地铁建设中所面临的主要技术难题,采用传统的室内试验评价湿陷性的方法具有一定的局限性。本文采用数值模拟方法分析了地铁工程基底湿陷性黄土应力特征,在此基础上,通过现场试坑浸水试验,准确判定了黄土场地湿陷类型及自重湿陷下限深度,研究了大厚度湿陷性黄土场地浸水过程中的自重湿陷变形特征、水分场的时空运移、土压力(竖向及侧向)、侧向位移的变化规律,查明了浸水影响范围及地面裂缝影响范围,确定了自重湿陷的地区土质修正系数β0,为湿陷性黄土地区地铁工程的地基处理提供依据。     

Neural network based prediction of ground surface settlements due to tunnelling

Ground surface settlement due to tunnel excavation varies in magnitude and trend depending on several factors such as tunnel geometry, ground conditions, etc. Although there are several empirical and semi-empirical formulae available for predicting ground surface settlement, most of these do not simultaneously take into consideration all the relevant factors, resulting in inaccurate predictions. In this study, an artificial neural network (ANN) is incorporated with '113' of monitored field results to predict surface settlement for a tunnel site with prescribed conditions. To achieve this, a standard format (a protocol) for a database of monitored field data is first proposed and then used for sorting out a variety of monitored data sets available in KICT. Using the capabilities of pattern recognition and memorization of the ANN, an attempt is made to capture the rich physical characteristics smeared in the database and at the same time filter inherent noise in the monitored data. Here, an optimal neural network model is suggested through preliminary parametric studies. It is shown that preliminary studies for generating an optimal ANN under given training data sets are necessary because no analytical method for this purpose is available to date. In addition, this study introduces a concept of relative strength of effects (RSE) [Yang Y, Zhang Q. A heirarchical analysis for rock engineering using artificial neural networks. Rock Mechanics and Rock Engineering 1997; 30(4): 207–22] in sensitivity analysis for various major factors affecting the surface settlement in tunnelling. It is seen in some examples that the RSE rationally enables us to recognize the most significant factors of all the contributing factors. Two verification examples are undertaken with the trained ANN using the database created in this study. It is shown from the examples that the ANN has adequately recognized the characteristics of the monitored data sets retaining a generality for further prediction. It is believed that an ANN based hierarchical prediction procedure shown in this paper can be further employed in many kinds of geotechnical engineering problems with inherent uncertainties and imperfections.