Direct back analysis by the meshless local Petrov‐Galerkin method and Bayesian statistics

This study presents a preliminary development of a direct back analysis procedure by the meshless local Petrov–Galerkin (MLPG) method and Bayesian statistics and the application of resulting procedure to characterize soil properties using laboratory results. As compared to direct back analyses that are based on the finite element method (FEM), it is intended to show that MLPG‐based direct back analyses may be more suitable for some kinds of characterization problems; for example, involving a complex subsurface stratification or the characterization of soil properties of just an inclusion of a soil profile. The existing MLPG method is first slightly modified to analyse time‐dependent problems. Using the resulting method, quantities to be characterized are evaluated so that they give numerical results as close to measured data as possible. The Akaike information criterion is introduced for simplifying the evaluation. A one‐dimensional finite strain consolidation problem is introduced to do an error analysis for prediction by the proposed MLPG method. Another example illustrates experiences of performing an MLPG‐based direct back analysis. Comparison of MLPG‐based and FEM‐based direct back analyses is taken. Copyright © 2006 John Wiley & Sons, Ltd.     

Characterizing soil properties by the meshless local Petrov-Galerkin method

This study presents a preliminary development of a direct back analysis procedure by the meshless local Petrov-Galerkin (MLPG) method and its application to characterize soil properties using in-situ test results. As compared to a direct back analysis based on the finite element method (FEM), it is intended to show that doing a direct back analysis by the MLPG method may reduce the computation costs or treat with the non-homogeneity of characterized soil properties more easily. Taking a two-dimensional (2D) solute transport problem as an illustration, an MLPG1 model of such a problem is derived to predict the solute concentration with trial coefficients of dispersion. To account for the non-homogeneity of these coefficients, variation of them is described by the moving least-squares interpolation. The Akaike information criterion is then introduced to find those coefficients of dispersion with which errors between predicted and measured data are minimized. A benchmark problem is studied to test the precision of numerical results provided by the proposed MLPG1 model. Another example illustrates the experiences of doing an MLPG-based direct back analysis. Comparison of MLPG-based and FEM-based direct back analyses is taken.     

基于人工蜂群算法的岩体结构面多参数优势分组研究

岩体结构面优势分组是研究岩体力学性质与水力特性的基础,通常的分析方法是只根据产状进行划分。鉴于结构面其他特征对岩体力学性质的重要影响,考虑结构面倾向、倾角、迹长、张开度、表面形态5个特征参数,提出了基于人工蜂群算法的岩体结构面多参数优势分组方法。以样本总体离差平方和为目标函数,建立岩体结构面多参数优势分组的数学模型,应用人工蜂群优化算法求解,以目标函数值最小时的解作为聚类中心,并自动确定分组边界。对人工生成的结构面数据的计算结果验证了该方法的正确性,该方法的求解精度是令人满意的。最后,将该方法应用于怒江松塔水电站坝址区岩体结构面多参数优势组的划分,得到了较为合理的分组结果,进一步验证了此方法具有较高的运行效率与工程实用性。     

Numerical characterization of dispersion coefficients by meshless‐based inverse analysis

This study presents the preliminary development of a meshless‐based indirect inverse analysis. Two goals are desired to be attained: (a) identifying fewer quantities in an indirect inverse analysis and (b) getting the interpolation results of spatial variation of a quantity insensitive to the replacement of one or few interpolation points with other points. This idea is illustrated by modifying a meshless local Petrov‐Galerkin (MLPG) formulation of a two‐dimensional (2D) contaminant transport problem under the concept of boundary control. Thus, dispersion coefficients to be identified are solved from measured or prescribed pollutant concentrations. A field problem is introduced to test the performance of resulting meshless‐based indirect inverse analysis procedure. This test finds a major motive and a minor one for performing a meshless‐based indirect inverse analysis. The major motive is attaining the aforementioned Goal (b), while the above‐mentioned Goal (a) is the minor motive. In conclusion, the development of a meshless‐based indirect inverse analysis procedure can be a valuable application of the MLPG method. Copyright © 2010 John Wiley & Sons, Ltd.     

高填方边坡失稳时间预测的实用模型

建立一种准确可靠的方法来预测高填方边坡因蠕变破坏而发生滑坡的时间是困难的,但对防止财产和生命损失又至关重要。在总结高填方土质边坡蠕变破坏过程中的位移、速度特征的基础上,通过改进Saito模型的应变率公式,提出了基于改进人工蜂群算法的滑坡中短期预测的实用模型。将进入加速变形阶段后的滑坡位移时间序列作为输入,通过人工蜂群算法反演实用模型参数后输出预测的滑坡时间。以3个高填方滑坡为实例,应用滑坡位移监测点的测量数据,验证了该方法在滑坡时间预测上的准确性和可靠性。同时,将该方法预测的滑坡时间结果与传统的Saito系列模型预测的滑坡时间结果进行了比较。结果表明,在通过滑坡位移的时间序列进行滑坡时间预测时,所提出的实用模型比两种Saito模型更准确可靠。     

基于时间序列与人工蜂群支持向量机的滑坡位移预测研究

总结以往滑坡预测方法存在的诸多不足,针对滑坡监测位移-时间曲线特点,本文提出了一种基于时间序列的人工蜂群算法（ABC）与支持向量回归机（SVR）相结合的滑坡位移预测方法。以三峡库区白水河滑坡为例,通过对滑坡位移、降雨、库水位等因素的分析,研究影响滑坡位移变化的因素。用时间序列加法模型和移动平均法将滑坡位移分解为趋势项和周期项。以多项式最小二乘法拟合滑坡位移趋势项,用人工蜂群支持向量机模型对滑坡位移周期项进行训练和预测。通过灰色系统关联分析法计算多项因子与滑坡位移周期项之间的关联性。最终的滑坡总位移预测值为周期项预测值与趋势项预测值之和。与BP神经网络、PSO-SVR模型方法相比,该方法在滑坡位移预测中有更高的精度,在防灾减灾工作中有较好的推广应用前景。     

基于ABC-BP模型的煤层含气量预测

煤层含气量预测是煤层气资源勘探开发利用前期的重要研究内容之一。近些年,BP神经网络算法常用于煤层含气量预测领域,但传统BP模型在训练过程中往往存在收敛速度慢、对初始值敏感以及易陷入局部极小值等问题。为此,提出了一种改进的以人工蜂群算法为特征的BP神经网络预测方法。以沁水盆地某工区3号煤层为研究对象,首先,利用R型聚类分析法对目标煤储层所提取的多种类型的地震属性进行分类,优选出4种对煤层含气量变化反应最敏感且相互独立的地震属性;再利用人工蜂群算法(ABC)寻找BP神经网络的输入层与隐含层的最优连接权值和隐含层的最优阈值,构建具有鲁棒性的ABC-BP神经网络预测模型,并以井位置优选地震属性和含气量数据为样本训练该模型;最后,以整个工区目标储层的优选地震属性为输入,进行工区内煤层含气量的预测。预测结果与各井含气量的变化趋势基本吻合,其中,训练井处的平均误差率为0.23%,验证井处的误差率低于15%,预测精度较高,因此,该预测方法可靠性高,适用性强,可有效用于煤层含气量预测。      

无网格局部Petrov-Galerkin法大地电磁二维正演

将无网格局部Petrov-Galerkin算法用于大地电磁二维正演。介绍了该方法的基本原理;从大地电磁二维边值问题出发,利用子域法详细推导了与之对应的局部Petrov-Galerkin弱式方程,并用高斯积分法将其离散化。论述了无网格局部Petrov-Galerkin法较无单元Galerkin法及有限元法的优缺点,最后通过二维模型的计算验证了算法的有效性。      

基于ABC-GRNN组合模型的露天矿边坡变形预测

准确预测露天矿边坡变形是有效实现边坡临灾预警的重要保证,针对传统边坡变形预测方法无法表征和综合分析边坡变形受多种因素的影响,提出一种露天矿边坡变形的人工蜂群(ABC)算法优化广义回归网络(GRNN)组合预测模型(ABC-GRNN)。在此预测模型中,综合考虑了影响露天矿边坡变形的5个因素：开采扰动、降雨量、降雨持续时间、温度以及湿度。以山西中煤平朔安家岭露天矿为例,通过遗传算法改进BP神经网络(GA-BPNN)、支持向量机(SVM)等人工智能算法与实测变形数据进行预测效果对比分析。结果表明：ABC算法能够快速帮助GRNN寻优获取合适的传递参数,并对变形进行有效的预测。ABC-GRNN组合预测模型,将预测结果的平均绝对误差292.9 mm、平均绝对百分比误差0.691 3%及均方根误差338.9 mm分别降低到25 mm、0.043 3%和29.5 mm,说明该模型具有更高的预测精度;ABC-GRNN模型比其他模型收敛速度快,只经过7步的迭代,即可得到最小的均方误差。与其他预测模型相比较,本文模型的预测精度更高、泛化能力更强、收敛速度更快,有较高的实用价值。     

Optimization techniques for identifying soil parameters in geotechnical engineering: Comparative study and enhancement

A comparative study of optimization techniques for identifying soil parameters in geotechnical engineering was first presented. The identification methodology with its 3 main parts, error function, search strategy, and identification procedure, was introduced and summarized. Then, current optimization methods were reviewed and classified into 3 categories with an introduction to their basic principles and applications in geotechnical engineering. A comparative study on the identification of model parameters from a synthetic pressuremeter and an excavation tests was then performed by using 5 among the mostly common optimization methods, including genetic algorithms, particle swarm optimization, simulated annealing, the differential evolution algorithm and the artificial bee colony algorithm. The results demonstrated that the differential evolution had the strongest search ability but the slowest convergence speed. All the selected methods could reach approximate solutions with very small objective errors, but these solutions were different from the preset parameters. To improve the identification performance, an enhanced algorithm was developed by implementing the Nelder‐Mead simplex method in a differential algorithm to accelerate the convergence speed with strong reliable search ability. The performance of the enhanced optimization algorithm was finally highlighted by identifying the Mohr‐Coulomb parameters from the 2 same synthetic cases and from 2 real pressuremeter tests in sand, and ANICREEP parameters from 2 real pressuremeter tests in soft clay.     

Three‐dimensional simulations of tensile cracks in geomaterials by coupling meshless and finite element method

Failure in geotechnical engineering is often related to tension‐induced cracking in geomaterials. In this paper, a coupled meshless method and FEM is developed to analyze the problem of three‐dimensional cracking. The radial point interpolation method (RPIM) is used to model cracks in the smeared crack framework with an isotropic damage model. The identification of the meshless region is based on the stress state computed by FEM, and the adaptive coupling of RPIM and FEM is achieved by a direct algorithm. Mesh‐bias dependency, which poses difficulties in FEM‐based cracking simulations, is circumvented by a crack tracking algorithm. The performance of our scheme is demonstrated by two numerical examples, that is, the four‐point bending test on concrete beam and the surface cracks caused by tunnel excavation. Copyright © 2014 John Wiley & Sons, Ltd.     

岩土工程数值计算中的无网格方法及其全自动布点技术

自然单元法采用无网格的思想全域构造插值函数,它的求解精度高,计算时间少,可准确地施加边界条件,兼具有无网格法和有限单元法的优点和特点,是一种理想的用于岩土及地下工程分析计算的数值方法。文中简要地介绍了自然单元法的基本理论,并针对岩土及地下工程问题特点,给出了一种无网格离散点的全自动布置方法。     

用最小二乘配点法解地下水稳定流问题

最小二乘配点无网格法是一种新型高效的无网格法。该方法除节点外又在研究域内引入辅助点,近似函数仍然只通过节点构造,微分方程在所有节点和辅助点上满足。本文将最小二乘配点无网格法应用于非均质多孔介质中的二维地下水稳定流问题,推导了计算格式、编制了相应的计算程序。算例结果表明,最小二乘配点无网格法算法简单,有较高的精度且节省计算量。     

自然单元法在岩土工程中应用两例

自然单元法(NEM)是较近出现的一种无网格方法,其形函数兼有无网格的特点和传统有限元的优点,是一种理想的适合岩土工程问题计算的新型数值方法。介绍了自然单元法的基本原理和特性,并讨论了其在岩土工程中的具体应用。将Goodman单元引入自然单元法以实现对不连续面的模拟,研究表明,在NEM中加入节理单元的总体原则和具体的实施细节与FEM中完全相同;而在一般的无网格方法中,则稍微复杂一点。为了实现对岩土工程中常见的无限域或半无限域问题的模拟,引入了无界单元;由于自然单元法的特性,自然单元法和无界元可实现无缝“耦合”。具体的数值算例验证了上述思路。     

A novel study of artificial bee colony with clustering technique on paddy rice image classification

Rice is a crop of global importance. To predict the area of paddy rice and thus its production, it draws great attraction of using data mining approaches on remote sensing data, which are well accepted. Many approaches based on supervised and unsupervised learning techniques have been developed over the years. Artificial bee colony (ABC) algorithm with a clustering technique is one of the most popular swarm-based algorithms. In this study, ABC algorithm is used to perform the rice image classification based on remote sensing imagery. This study comprises two stages. In the first part of the study, the ancillary information composed from the original spectra is applied to increase the performance of classification. As the other parts of the study, an efficient unsupervised classifier is developed to evaluate the performance of the incorporated ancillary information. This study integrates the ABC algorithm into a clustering process to build a land cover classifier system. On the other hand, a parallel approach using ant colony optimization (ACO) is studied for comparison. Two significant contributions are presented in this study: (1) a paddy rice image classifier is built with ABC algorithm and (2) the outcome of classifier using ABC algorithm outperforms that using ACO algorithm.     

基于国产高分辨率光学遥感影像的水体提取

遥感图像中地表水体同山体、建筑物等地物产生的阴影在光谱特征上存在较高的类间相似性,导致提取过程中容易出现混淆和错分的情况。针对此问题,提出一种基于面向对象和人工蜂群的地表水体提取方法。该方法首先对遥感图像进行分割以获取分割对象的光谱、比率、几何形状等统计特征,以弥补高分遥感图像波段数目少,信息量不足的缺陷;并借助人工蜂群算法在解决复杂问题最优化方面的优势,选取水体同阴影二值分类的几何平均正确率作为算法的适应度函数,最终获取地表水体的最优化提取规则。选取厦门市大嶝岛和湖南省资兴市部分区域,基于国产高分一号、二号遥感数据进行水体提取,并与传统SVM分类结果进行比较。实验结果表明本算法提取水体的总体精度和Kappa系数均优于传统SVM分类器,表明该方法可应用于高分遥感图像的地表水体提取。     

无网格自然单元法在弹塑性分析中的应用

无网格自然单元法在构造位移插值函数时不需要单元的信息,只需要结点的信息,在裂纹扩展模拟、材料非线性分析、几何非线性分析以及三维计算等方面具有广阔的应用前景。阐述了将无网格自然单元法应用于结构弹塑性分析的过程和基本理论,给出了其位移插值函数的构造过程,并将其结果与常规的有限单元法进行对比,证明了该方法用于弹塑性分析的优越性。     

Predicting longitudinal dispersion coefficient using ANN with metaheuristic training algorithms

A reliable prediction of dispersion coefficient can provide valuable information for environmental scientists and river engineers as well. The main objective of this study is to apply intelligence techniques for predicting longitudinal dispersion coefficient in rivers. In this regard, artificial neural network (ANN) models were developed. Four different metaheuristic algorithms including genetic algorithm (GA), imperialist competitive algorithm (ICA), bee algorithm (BA) and cuckoo search (CS) algorithm were employed to train the ANN models. The results obtained through the optimization algorithms were compared with the Levenberg–Marquardt (LM) algorithm (conventional algorithm for training ANN). Overall, a relatively high correlation between measured and predicted values of dispersion coefficient was observed when the ANN models trained with the optimization algorithms. This study demonstrates that the metaheuristic algorithms can be successfully applied to make an improvement on the performance of the conventional ANN models. Also, the CS, ICA and BA algorithms remarkably outperform the GA and LM algorithms to train the ANN model. The results show superiority of the performance of the proposed model over the previous equations in terms of DR, R ² and RMSE.     

Neural network prediction of the reliability of heterogeneous cohesive slopes

The reliability of heterogeneous slopes can be evaluated using a wide range of available probabilistic methods. One of these methods is the random finite element method (RFEM), which combines random field theory with the non‐linear elasto‐plastic finite element slope stability analysis method. The RFEM computes the probability of failure of a slope using the Monte Carlo simulation process. The major drawback of this approach is the intensive computational time required, mainly due to the finite element analysis and the Monte Carlo simulation process. Therefore, a simplified model or solution, which can bypass the computationally intensive and time‐consuming numerical analyses, is desirable. The present study investigates the feasibility of using artificial neural networks (ANNs) to develop such a simplified model. ANNs are well known for their strong capability in mapping the input and output relationship of complex non‐linear systems. The RFEM is used to generate possible solutions and to establish a large database that is used to develop and verify the ANN model. In this paper, multi‐layer perceptrons, which are trained with the back‐propagation algorithm, are used. The results of various performance measures indicate that the developed ANN model has a high degree of accuracy in predicting the reliability of heterogeneous slopes. The developed ANN model is then transformed into relatively simple formulae for direct application in practice. Copyright © 2016 John Wiley & Sons, Ltd.     

Determination of the critical failure surface for slope stability analysis using ant colony optimization

Slope stability analysis of any natural or artificial slope aims at determining the factor of safety of the slip surface that possesses the lowest factor of safety. In this study, an ant colony optimization (ACO) algorithm is developed to solve this factor-of-safety minimization problem. Factors of safety of slip surfaces are found by using the Morgenstern–Price method, which satisfies both force and moment equilibrium. Nonlinear equations from the Morgenstern–Price method are solved numerically by the Newton–Raphson method. In the proposed ACO algorithm, the initiation point and the shape of the slip surface are treated as the search variables. The proposed heuristic algorithm represents slip surfaces as piecewise-linear curves and solves for the optimal curve yielding the minimum factor of safety. To demonstrate its applicability and to investigate the validity and effectiveness of the algorithm, four examples with varying complexity are presented. The obtained results are compared with the available literature and are found to be in agreement.