二维场地液化势预测的神经网络方法

基于人工神经网络,提出了场地液化势预测模型。场地液化势的空间数据结构特征可由不同参数的自回归神经网络（GRNN）来模拟。该预测模型的一个重要参数spread可用地质统计学（Kriging）方法中的交叉验证技术来确定。研究表明,在最优spread参数条件下GRNN能够较好地映射场地液化势数据结构特征。用GRNN模型预测结果与经典的Kriging估计方法所得到的结果十分吻合。GRNN模型可以用于二维空间数据的预测及基于GIS决策系统。     

空间递推平均法在估计土性指标自相关距离中的应用

像许多土层一样,黏性土层的土性参数也具有空间变异性。土性参数的空间变异性主要是由沉积和成岩等地质作用和环境因素导致的,而相关距离的确定是分析土性参数空间变异性的关键。为了分析天津南疆港地区土层参数的空间变异性,对确定土性参数相关距离的空间递推平均法及其改进方法进行了汇总分析,并结合地质勘查资料,运用改进的空间递推平均法对天津港地区典型土层的垂直相关距离进行了统计分析,得到了典型土层地区性代表值,对该地区岩土工程的可靠度分析及勘探点位的选取具有重要的指导意义。此外,还结合工程实例讨论了相关距离的贝叶斯估值法,贝叶斯估值法比直接取平均值确定自相关距离的方法更为可靠、合理,应用贝叶斯定理对天津港南疆石化码头淤泥及淤泥质黏土层的自相关距离进行估值,为今后该地区的取样间距选择提供参考,也为可靠度理论在该地区岩土工程中的应用提供有价值的参数。     

基于Fredlund粒度分布曲线方程的中国黄土粒度分析

粒度分布曲线是土壤最基本的土性参数之一,通过数学方程预测粒度分布曲线将为工程勘察节省大量成本。Fred? lund建立在Fredlund和Xing土水特征曲线方程基础上的粒度分布曲线方程已被证明适用于多种土类,但其对中国黄土的有 效性很少得到验证。本文采用Fredlund粒度分布曲线方程对黄土高原7个地区18个黄土样的粒度分布曲线进行了拟合,结 果表明Fredlund粒度分布曲线方程与中国黄土的粒度分布曲线拟合度高,拟合参数稳定且呈规律性变化,能够很好地反映 中国黄土的地域性。本文得到的参数可用于各个地区的粒度分布曲线预测,并对黄土粒径分析及分类具有指导意义。     

吕梁山区黄土物理性质力学参数区域分布特征

黄土因其特殊的物理力学性质及工程特性,在外界环境影响下易引发诸如黄土崩塌、滑坡、泥流等地质灾害。因此,研究黄土土性参数区域性分布特征对地质灾害防灾减灾具有重要意义。以吕梁山区L₁黄土地层为研究对象,在野外调查、代表性点位取样的基础上,对研究区内133个取样点340组黄土试样进行黄土土性参数室内试验,获得其土性参数;基于所得数据,通过统计分析及ArcGIS软件平台,分析研究区内L₁黄土地层土性参数分布特征。结果表明:黏粒含量、天然含水率、孔隙比、黏聚力、内摩擦角的分布在南北方向上具有较好的规律性,山脉东西两侧规律略有不同。另外,黏粒含量、天然含水率、黏聚力、湿陷系数的区域分布规律较为明显,内摩擦角区域分布上较为离散。研究结果为从土性特征方面对地质灾害易发性评价提供基础数据支撑。     

苏中腹地湖相软土土性参数变异性统计描述

基于江苏中部某高速公路工程地质勘察所提供的大量室内试验测试数据,运用随机场理论,将土层剖面模拟为随机场而非传统意义上的随机变量,提出了基于随机场理论、考虑空间趋势分量的土性参数变异性统计方法,系统地统计分析了苏中腹地湖相软土土性参数的变异特性,对比了实验室物理性质指标参数、变形参数和强度参数变异性特征。统计结果表明,该地区软土实验室物理性质指标参数变异性较小,不存在明显的趋势分量,可用传统方法进行统计估计;实验室变形和强度参数存在明显的趋势分量,统计前需对数据进行回归分析和去趋势化处理,从而基于随机场理论进行统计分析。研究不仅提供了土性参数变异性统计的新方法,而且能对苏中地区工程建设提供必要的参考,为区域性软土土性参数随机场模型的建立打下了坚实的基础。     

西安黄土土性相关距离的统计分布特性研究

土性指标概率模型参数的确定是进行岩土工程可靠度分析和设计的基础。为了更好地推行岩土工程可靠度设计,应该首先建立地区土性参数的概率模型。作为反映土性指标自相关特性的相关距离,研究其区域分布特性,建立其概率分布模型是必要的。本文收集了西安黄土的176个钻孔CPT数据,利用其端阻值qc作为样本,采用递推空间法计算了各层西安黄土的相关距离。研究了计算结果的统计特性,并对其均值进行了单侧置信区间估计,提出了各层西安黄土相关距离的代表值。建立了西安各层黄土相关距离的概率模型,对模型参数进行了估计,采用皮尔逊-卡方检验法对模型进行了拟合优度检验,检验结果认为其符合Beta分布。     

西安黄土土性相关距离的统计分布特性研究

土性指标概率模型参数的确定是进行岩土工程可靠度分析和设计的基础。为了更好地推行岩土工程可靠度设计,应该首先建立地区土性参数的概率模型。作为反映土性指标自相关特性的相关距离,研究其区域分布特性,建立其概率分布模型是必要的。本文收集了西安黄土的176个钻孔CPT数据,利用其端阻值q c作为样本,采用递推空间法计算了各层西安黄土的相关距离。研究了计算结果的统计特性,并对其均值进行了单侧置信区间估计,提出了各层西安黄土相关距离的代表值。建立了西安各层黄土相关距离的概率模型,对模型参数进行了估计,采用皮尔逊-卡方检验法对模型进行了拟合优度检验,检验结果认为其符合Beta分布。     

天津港土性相关距离的计算研究和统计分析

相关距离是随机场理论应用于实际工程可靠度分析的一个重要参数。根据随机场理论,对土性参数相关距离的计算方法进行了研究和改进,提出求解相关距离的波动函数法和加权拟合相关函数法,使相关距离的计算更加简便易行,且精度有所提高。结合大量工程地质勘察资料,对天津港地区典型土层的垂直向相关距离及水平向相关距离进行了计算和统计,获得相关距离的地区性代表值,可作为此地区实际工程的参考。同时,对取样间距及土性指标对相关距离的影响进行了分析,提出了当取样间距变化时相关距离的确定原则,并指出相关距离是反映土的空间固有变异性的基本属性,由不同土性指标求得的相关距离值基本相同。     

黄土土性参数的统计分析

讨论了土性参数不确定性的主要来源, 并对黄土高原典型地段黄土土性参数进行了详细地统计分析, 结果表明黄土土性参数的均值、标准差和变异系数具有明显的区域性特点, 土性参数的统计必须针对具体的工程进行。黄土的物理力学指标的分布多服从正态分布, 在边坡可靠度计算中各指标均可用正态分布近似。     

相关函数法计算相关距离的分析探讨

根据数百个土层土性参数相关距离的计算, 对相关函数法计算相关距离的方法、相关函数型式选择、拟合范围等问题进行了分析探讨。     

Comparison of three artificial neural network approaches for estimating of slake durability index

Slake durability index (I _d2) is an important engineering parameter to assess the resistance of clay-bearing and weak rocks to erosion and degradation. Standard test sample preparation for slake durability test is difficult for some rock types and the test is time-consuming. The paper reports an attempt to define I _d2 using other parameters that are simpler to obtain. In this study, three different artificial neural network approaches, namely feed-forward back propagation (FFBP), radial basis function based neural network (RBNN), and generalized regression neural networks (GRNN) were used for estimating I _d2. The determination coefficient (R ²), root mean square error and mean absolute relative error statistics were used as evaluation criteria of the FFBP, RBNN, and GRNN models. The experimental results were compared with these models. The comparison results indicate that the GRNN models are superior to the FFBP and RBNN models in modeling of the slake durability index (I _d2).     

Reliability Analysis of Quick Sand Condition

First order reliability method (FORM) is generally used for reliability analysis in geotechnical engineering. This article adopts generalized regression neural network (GRNN) based FORM, Gaussian process regression (GPR) based FORM and multivariate adaptive regression spline (MARS) based FORM for reliability analysis of quick sand condition. GRNN is related to the radial basis function (RBF) network. GPR is developed based on probabilistic framework. MARS is a nonparametric regression technique. A comparative study has been carried out between the developed models. The performance of GPR based FORM and MARS based FORM match well with the FORM. This article gives the alternative methods for reliability analysis of quick sand condition.     

Investigation of general regression neural network architecture for grade estimation of an Indian iron ore deposit

The mineral resource estimation requires accurate prediction of the grade at location from limited borehole information. It plays the dominant role in the decision-making process for investment and development of various mining projects and hence become an important and crucial stage. This paper evaluvates the use of two distinct artificial neural network (ANN)-based models, general regression neural network (GRNN) and multilayer perceptron neural network (MLP NN), to improve the grade estimation from Koira iron ore region in Sundargarh district, Odisha. ANN-based models capture the inherent complex structure of mineral deposits and provide a reliable generalization of the iron grade. The ANN-based approach does not require any preliminary geological study and is free from any statistical assumption on the raw data before its application. The GRNN is a one-pass learning algorithm and does not require any iterative procedure for training less complex structure and requires only one learning parameter for optimization. In this investigation, the spatial coordinates and multiple lithological units were taken as input variables and the iron grade was taken as the output variable. The comparative analysis of these models has been carried out and the results obtained were validated with traditional geostatistical method ordinary kriging (OK). The GRNN model outperforms the other methods, i.e. MLP and OK, with respect to generalization and predictability of the grades at an un-sampled location.     

基于遗传－广义回归神经元算法的坞石隧道三维弹塑性位移反分析研究

广义回归神经元网络在逼近能力、学习速度和网络稳定性方面均优于BP神经元网络,且具有网络人为调节参数少的优点。本文将广义回归神经元网络引入坞石隧道工程的三维弹塑性位移反分析。为了在网络训练过程中快速搜索到最优的网络阈值,采用十进制遗传算法对网络阈值进行优化。在确定最优的网络结构后,采用遗传算法在每个待反演参数的搜索范围内搜索出与实测位移最接近的围岩力学与初始应力场参数组合。用反分析得来的参数进行下步开挖位移预测,预测值与实测值吻合较好,表明所提出的这种反分析方法在工程上是可行的,可以推广使用。     

Simulation of the availability index of soil copper content using general regression neural network

Excessive soil copper (Cu) availability leads to plant growth retardation and leaf chlorosis, and the contamination of Cu in the food chain would be detrimental to human and animal health. The most important path for Cu accumulation in plants is uptake from soils. It is therefore important to understand the availability of soil Cu and its controlling factors to modify Cu availability and prevent excessive Cu from entering the food chain. The present study proposed a general regression neural network (GRNN) to simulate the availability index of soil Cu (available heavy mental concentrations/total heavy metal concentrations), based on the influencing factors of total Cu concentration, pH, organic matter (OM), available phosphorus (AP), and readily available potassium (RAK). Results showed that total Cu concentration, combined with OM and AP, achieved the lowest RMSE value (0.0524) for the modeled value of the availability index of soil Cu. The simulated results by GRNN and the ground truth values had better agreement (R ² = 0.7760) than that by a linear model (R ² = 0.6464) for 23 test samples. Moreover, GRNN obtained lower averaged relative errors than linear model. This demonstrated that GRNN could be used to simulate the availability index of soil heavy metals and gained better results than linear model.     

Specification and prediction of nickel mobilization using artificial intelligence methods

Groundwater and soil pollution from pyrite oxidation, acid mine drainage generation, and release and transport of toxic metals are common environmental problems associated with the mining industry. Nickel is one toxic metal considered to be a key pollutant in some mining setting; to date, its formation mechanism has not yet been fully evaluated. The goals of this study are 1) to describe the process of nickel mobilization in waste dumps by introducing a novel conceptual model, and 2) to predict nickel concentration using two algorithms, namely the support vector machine (SVM) and the general regression neural network (GRNN). The results obtained from this study have shown that considerable amount of nickel concentration can be arrived into the water flow system during the oxidation of pyrite and subsequent Acid Drainage (AMD) generation. It was concluded that pyrite, water, and oxygen are the most important factors for nickel pollution generation while pH condition, SO₄, HCO₃, TDS, EC, Mg, Fe, Zn, and Cu are measured quantities playing significant role in nickel mobilization. SVM and GRNN have predicted nickel concentration with a high degree of accuracy. Hence, SVM and GRNN can be considered as appropriate tools for environmental risk assessment.     

基于EEMD-PSO-GRNN模型的元江中上游径流预测

径流序列的非线性和非平稳特性使得高精度的径流预报存在困难。本文组合EEMD和GRNN模型形成EEMD-GRNN耦合模型,预测时通过将径流序列分解为确定成分与随机成分并通过GRNN模型分别进行预测,预测值的加和则构成径流最终预测结果。EEMD-GRNN耦合模型应用到元江中上游,并与其他模型进行比较,结果表明：EEMD-GRNN耦合模型具有更高的预测精度,对径流的总体趋势预测有良好的效果,但在随机性的模拟上有待进一步完善。EEMD-GRNN耦合模型优于BP、GRNN、EEMD-BP模型,能有效提升径流预测的精度,可为流域的水资源优化调度等提供决策支持。     

致密砂岩气层压裂产能及等级预测方法

致密砂岩储层孔隙度小、渗透率低、含气饱和度低,基本上没有自然产能,需要进行压裂,因此进行压裂产能的预测很有必要。笔者研究了鄂尔多斯盆地苏里格东部地区盒8段致密砂岩气层的压裂产能及等级预测。利用反映储层流动性质的测井参数(电阻率、自然伽马、声波时差、中子、密度)与反应压裂施工情况的压裂施工参数(单位厚度砂体积、砂比、砂质量浓度、单位厚度排量、单位厚度入井总液量),选择数学统计方法神经网络法进行致密砂岩气层压裂产能等级预测。分析比较Elman神经网络、支持向量回归(SVR)、广义回归神经网络(GRNN)3个神经网络预测致密砂岩气层压裂产能模型的网络结构参数、回判及预测精度以及运行耗费时间。结果表明:3个模型中,GRNN网络参数只有1个,回判和预测精度最高,运行时间小于1 s。因此,选择GRNN模型预测致密砂岩气层压裂产能,并用于苏里格东部地区致密砂岩气层压裂产能的等级预测。等级预测准确率达到90%。     

直流电阻率测深二维自适应正则化反演

直流电阻率测深二维反演中,正则化参数的选取影响反演结果分辨率及反演过程稳定性。利用主动约束平衡正则化因子,进行直流电阻率光滑约束最小二乘二维自适应反演,改善直流电阻率测深二维反演的分辨率与稳定性。在反演迭代过程中,正则化因子根据模型参数的空间展布函数进行自适应计算、正则化参数的自适应计算。模拟数据反演结果验证了该方法的有效性与可行性,反演结果能准确地反映地下模型的真实电性结构。     

Flexible spectral methods for the generation of random fields with power-law semivariograms

Random field generators serve as a tool to model heterogeneous media for applications in hydrocarbon recovery and groundwater flow. Random fields with a power-law variogram structure, also termed fractional Brownian motion (fBm) fields, are of interest to study scale dependent heterogeneity effects on one-phase and two-phase flow. We show that such fields generated by the spectral method and the Inverse Fast Fourier Transform (IFFT) have an incorrect variogram structure and variance. To illustrate this we derive the prefactor of the fBm spectral density function, which is required to generate the fBm fields. We propose a new method to generate fBm fields that introduces weighting functions into the spectral method. It leads to a flexible and efficient algorithm. The flexibility permits an optimal choice of summation points (that is points in frequency space at which the weighting function is calculated) specific for the autocovariance structure of the field. As an illustration of the method, comparisons between estimated and expected statistics of fields with an exponential variogram and of fBm fields are presented. For power-law semivariograms, the proposed spectral method with a cylindrical distribution of the summation points gives optimal results.