Ensemble-based hierarchical multi-objective production optimization of smart wells

In an earlier study, two hierarchical multi-objective methods were suggested to include short-term targets in life-cycle production optimization. However, this earlier study has two limitations: (1) the adjoint formulation is used to obtain gradient information, requiring simulator source code access and an extensive implementation effort, and (2) one of the two proposed methods relies on the Hessian matrix which is obtained by a computationally expensive method. In order to overcome the first of these limitations, we used ensemble-based optimization (EnOpt). EnOpt does not require source code access and is relatively easy to implement. To address the second limitation, we used the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm to obtain an approximation of the Hessian matrix. We performed experiments in which water flood was optimized in a geologically realistic multilayer sector model. The controls were inflow control valve settings at predefined time intervals. Undiscounted net present value (NPV) and highly discounted NPV were the long-term and short-term objective functions used. We obtained an increase of approximately 14 % in the secondary objective for a decrease of only 0.2–0.5 % in the primary objective. The study demonstrates that ensemble-based hierarchical multi-objective optimization can achieve results of practical value in a computationally efficient manner.     

考虑水库岸坡特征的滑坡稳定性计算修正传递系数法

随着低碳能源的推行,大量水电站的兴建引发了大量库岸滑坡问题。库岸滑坡与普通滑坡有着显著区别,但在实际工程的稳定性计算过程中,工程人员常将二者归为一体,采用同样的思路和方法进行研究。为了更好地治理库岸滑坡,根据库岸自身的特点,在长期野外地质调查和室内文献调研的基础上,提出按结构面特征、坡体位置和滑体物质成分进行库岸斜坡分区的原则;针对库岸滑坡的特点,推导了被牵引段有推力+后缘拉裂面有剪力、被牵引段不存在推力+后缘拉裂面有剪力、被牵引段不存在推力+后缘拉裂面无剪力三种情况的水库岸坡滑坡稳定性计算公式;结合工程实例,验证基于水库岸坡特征滑坡稳定性计算修正传递系数法的正确性。     

基于卷积神经网络迁移学习模型的矿岩智能识别方法

文章基于Inception-v3卷积神经网络模型,通过对采集的金矿石、铜矿石、铁矿石、铅锌矿、花岗岩、片麻岩、大理岩和页岩,8种岩石453张图像进行特征提取和迁移学习,建立了岩性分类的迁移学习模型,实现了岩性的自动识别和分类。每种岩石图像随机抽取4张作为测试集进行测试,剩余421张图像作为训练集参加训练,经测试全部图像的岩性分类结果均正确,识别正确率超过80%的岩石图像占测试集图像总数的90%以上。识别正确率未达到80%的图像经过处理后重新训练并测试,其识别正确率均超过了80%,表明了该模型具有良好的岩性识别能力且鲁棒性较好,为岩性识别和自动分类提供了一种新的智能分析方法。     

An adjoint-based optimization method for gas production in shale reservoirs

In this work, we consider a new model for flow in a multiporosity shale gas reservoir constructed within the framework of an upscaling procedure where hydraulic fractures are treated as (\(n-1\)) interfaces (\(n=2,3\)). Within this framework, the hydrodynamics is governed by a new pressure equation in the shale matrix which is treated as a homogenized porous medium composed of organic matter (kerogen aggregates with nanopores) and inorganic impermeable solid (clay, calcite, quartz) separated from each other by a network of interparticle pores of micrometer size. The solution of the pressure equation is strongly influenced by the constitutive response of the retardation parameter and effective hydraulic conductivity where the former incorporates gas adsorption/desorption in the nanopores of the kerogen. By focusing our analyses on this nonlinear diffusion equation in the domain occupied by the shale matrix, an optimization strategy seated on the adjoint sensitivity method is developed to minimize a cost functional related to gas production and net present value in a single hydraulic fracture. The gradient of the objective functional computed with the adjoint formulation is explored to update the controlled pressure drop aiming to optimize production in a given window of time. The combination of the direct approach and gradient-based optimization using the adjoint formulation leads to the construction of optimal production scenarios under controlled pressure decline in the well. Numerical simulations illustrate the potential of the methodology proposed herein in optimizing gas production.     

地下水污染监测网多目标优化设计模型及进化求解

采用模拟-优化方法建立了一个用于地下水污染监测网设计的多目标优化模型,该模型包括最小化监测费用、污染物质量评估误差、污染羽一阶矩评估误差和二阶矩评估误差等4个目标函数,以充分揭示减少地下水污染监测费用与提高污染监测精度之间的权衡关系.将改进小生境Pareto遗传算法与地下水流模拟程序和污染物运移模拟程序相耦合用于求解地下水污染监测网多目标设计模型.算例研究表明,采用进化算法求解监测网的多目标模型,能真实地反映各个目标函数间的权衡关系,并且不用考虑传统方法中惩罚因子的影响.与单目标优化模型相比,多目标优化模型可在较短的时间内得到优化问题的一系列Pareto权衡解,以利于相应条件下决策者选择最为经济有效的地下水污染监测方案.     

An intelligent response surface method for analyzing slope reliability based on Gaussian process regression

Problems in geotechnical engineering inevitably involve many uncertainties in the analysis. Reliability methods are important for evaluating slope stability and can take the uncertainties into consideration. In this paper, a novel intelligent response surface method is proposed in which a machine learning algorithm, namely Gaussian process regression, is used to approximate the high-dimensional and highly nonlinear response hypersurface. An iterative algorithm is also proposed for updating the response surface dynamically by adding the new training point nearest to the limit state surface to the initial training database at each step. The proposed Gaussian process response surface method is used to analyze three different case studies to assess its validity and efficiency. Direct Monte Carlo simulation is also carried out in each case to serve as the benchmark. Comparing with other methods confirms the accuracy and efficiency of the novel intelligent response surface method, which requires fewer performance function calls and avoids the need to normalize the correlative non-normal variables.     

隧洞围岩损失位移估计的智能优化反分析

隧洞开挖过程中围岩监测断面的布置一般滞后于掌子面开挖,监测断面布置前围岩已发生的位移称为损失位移。采用优化反分析思路求取损失位移,该思路将损失位移的求解转化为以实测位移与计算位移的误差作为目标函数、岩体力学参数作为决策变量的全局优化反分析问题。针对该全局优化反分析问题是一类高度非线性多峰值且计算代价较高的优化问题,将性能优异的粒子群优化算法与高斯过程机器学习方法相融合,结合FLAC3D数值计算程序,提出隧洞围岩损失位移优化反分析的粒子群-高斯过程-FLAC3D智能协同优化方法。算例研究表明,该方法是可行的,不仅能获得可靠的损失位移预测结果,而且可获取合理的围岩计算模型力学参数,具有全局性好、计算效率高的特点,克服了传统优化反分析方法容易陷入局部最优或过于依赖初始学习样本的局限性。将该方法应用到锦屏二级水电站辅助洞BK14+599断面的损失位移反分析,获得了该断面围岩的损失位移和力学参数,其中,损失位移较大,原因在于岩体开挖后在短时间内弹性变形大。因此,对于地下工程,特别是深部地下岩体工程,在围岩稳定性评价与围岩参数反分析中,损失位移不可忽视,应给予足够重视。     

An intelligent multi-objective EPR technique with multi-step model selection for correlations of soil properties

Current multi-objective evolutionary polynomial regression (EPR) methodology has difficulties on decision-making of optimal EPR model. This paper proposes an intelligent multi-objective optimization-based EPR technique with multi-step automatic model selection procedure. A newly developed multi-objective differential evolution algorithm (MODE) is adopted to improve the optimization performance. The proposed EPR process is composed of two stages: (1) intelligent roughing model selection and (2) model delicacy identification. In the first stage, besides two objectives (model accuracy and model complexity), the model robustness measured by robustness ratio is considered as an additional objective in the multi-objective optimization. In the second stage, a new indicator named selection index is proposed and incorporated to find the optimal model. After intelligent roughing selection and delicacy identification, the optimal EPR model is obtained considering the combined effects of correlation coefficient, size of polynomial terms, number of involved variables, robustness ratio and monotonicity. To show the practicality of the proposed EPR technique, three illustrative cases helpful for geotechnical design are presented: (a) modelling of compressibility, (b) modelling of undrained shear strength and (c) modelling of hydraulic conductivity. For each case, a practical formula with better performance in comparison with various existing empirical equations is finally provided. All results demonstrate that the proposed intelligent MODE-based EPR technique is efficient and effective.

     

An intelligent displacement back-analysis method for earth-rockfill dams

An intelligent method for the effective displacement back-analysis of earth-rockfill dams was proposed by combining artificial neural networks and evolutionary calculation. This method employs artificial neural networks, with optimal architecture trained by the evolutionary calculation and Vogl’s algorithm, instead of the time-consuming finite element analysis. In the back analysis, the soil parameters were optimized by performing evolutionary calculations on the tested neural network. The proposed method was verified by applying it to the displacement back-analysis of two projects in China, and the influence of generation number and set size on the simulation ability of neural networks was investigated.     

基于GIS的腊撒水电站库区渗漏地段初步分析

由于库区的广域性和地质条件的复杂性,水库渗漏问题一直是水利水电工程中主要工程地质问题.近年来遥感技术、GIS(Geographic Information System,地理信息系统)技术、三维技术的发展,为水库渗漏分析提供了一个有力的工具平台.本文以某水库库区渗漏问题为研究对象,基于遥感技术、GIS技术、三维技术建立的三维遥感影像可视化系统,对某库区地形数据进行分析,结合地层岩性、地质构造、水动力条件等,辨识出库区可能渗漏通道,进行渗漏估算,判断库区水库渗漏情况.     

Study on the improving method for gas production prediction in tight clastic reservoir

Relative permeability and resistivity index are important parameters for the exploration and development in a tight sandstone gas field. In the splitting method which uses permeability (K), reservoir thickness (H), and relative permeability (K), briefly referred to as the KHK splitting method, the accuracy of the relative permeability is crucial. According to the relationship between resistivity index and relative permeability of the Mesozoic Lower Safa Formation at Obaiyed Field in the Western Desert of Egypt, we improved the split method and made it more in line with the real situation by adopting Pairoys’ model which is more suitable to our study area. In this paper, we use a radial basis function (RBF) to establish the relationship between logging data and the gas production split point to point in tight sandstone gas reservoirs. To compare with the result by support vector regression (SVR), our method is better as indicated by mean absolute error values. In order to solve the problem that the relative permeability is difficult to obtain in the well logging evaluation, we also provide a convenient method and application example.     

基于可靠度理论的阻滑键加固渣土边坡多目标优化设计方法

随着城市建设的发展,渣土边坡的数量和规模急剧增加,渣土边坡的防控研究受到了广泛关注。针对渣土边坡人工分层堆填建筑余渣土体参数的不确定性,采用预埋阻滑键加固渣土边坡的方式,提出了基于可靠度理论的阻滑键多目标优化设计方法。考虑不同阻滑键潜在组合对渣土边坡预估破坏损失的影响,将渣土边坡的预估破坏损失、稳定安全性和加固设计成本作为设计目标,通过多目标优化理论确定帕累托前沿并计算其关节点,获得预埋阻滑键加固边坡的最佳设计方案。以深圳市某渣土边坡为例,计算结果表明,将破坏概率作为衡量阻滑键加固渣土边坡效果指标时,应在渣土边坡前缘预埋两组尺寸长3 m、间距5 m的阻滑键加固边坡。采用上述阻滑键设计组合加固该渣土边坡时,可实现该边坡预估破坏损失、设计成本和稳定性达到最佳平衡。     

基于SLR-ANN的地应力场三维智能反演方法研究

基于黄登引水发电系统区域地应力实测结果及建立的三维数值仿真计算模型,揭示了工程所在区域的三维地应力场分布特征,为地下工程的开挖加固设计提供更加准确的基础资料。分别采用传统多元线性回归方法、人工神经网络方法与考虑地质历史过程的基于逐步回归原理耦合人工神经网络(SLR-ANN)的非线性智能方法获得黄登水电站厂址区域的地应力场,再将地应力的实测值与反演数值解进行对比。结果表明:3种方法下反演所得引水发电系统区域内三维地应力场均与实测结果相一致,表明3种方法较为真实地模拟了整个地下洞室群区域三维地应力场的分布规律及特征。但采用SLR-ANN二次智能反演方法进行地应力反演,模拟效果更加接近监测值,且因减少了反演参数的个数而大幅度地提高了反演效率,可将反演计算结果应用于后续洞室开挖及锚固仿真分析中。     

基于多目标函数的黏弹性全波形反演

全波形反演是勘探地球物理领域兴起的核心技术之一,不但可以构建地下速度结构,也能够反演衰减参数(品质因子Q)模型,有助于识别地下介质类型和构造(如流体和煤层陷落柱),对煤和油气等自然资源的勘探和开发有重要意义。参数串扰是黏弹性全波形反演的关键难点,受速度误差影响,反演的Q模型会包含非常强的串扰噪声。针对该问题,提出了基于多目标函数的黏弹性全波形反演理论与方法,首先通过旅行时反演速度结构,再通过中心频率目标函数反演Q模型,最终使用波形差目标函数同时反演速度和Q模型。由于中心频率主要受衰减影响,因此,可有效减弱速度误差对Q反演的影响。最后,通过数值模拟验证了算法可有效地反演速度和Q模型。     

Groundwater risk assessment based on optimization framework using DRASTIC method

Due to anthropogenic influences and large amounts of pollutant released into the groundwater, it is vital to investigate groundwater quality and to characterize susceptible areas to contamination. In this paper, a new optimization-based methodology is proposed for determining groundwater risk using DRASTIC model based on genetic algorithm optimization model and Wilcoxon test. The correlation coefficient between DRASTIC/modified DRASTIC indices and nitrate concentrations in monitoring wells is used as a criteria for evaluating the efficiency of the proposed models. In this regard, because of the unsatisfactory original DRASTIC’s result, sensitivity analysis, genetic algorithm (GA), and Wilcoxon test (1945) are carried out to tackle the subjectivity associated with the original DRASTIC model and obtain better and reliable results. The results indicate that application of Wilcoxon test and GA optimization outperforms the others. Consequently, the correlation coefficient increased remarkably as compared to the original DRASTIC model (from 0.57 to 0.82). The proposed optimization process is adaptable to be applied in different case studies; mainly since it has the ability to optimize the weights of the model based on hydrogeological characteristics of the aquifer. Finally, the risk maps of the models are prepared using ArcGIS® to determine the most vulnerable areas.     

基于支持向量机的水库诱发地震分析

将诱发水库地震的主要因素（岩性、岩体完整性、断层性质、库区区域应力状态、库区地震活动背景）划分为11个因子,并进行定量化;再根据每个样本到所属类内超平面的距离计算每个样本点的模糊因子,确定其对分类超平面影响大小;然后建立水库地震的支持向量机（SVM）和模糊支持向量机（FSVM）模型,并应用于水库诱发地震等级预测。实例分析表明,两种模型均可用于水库诱发地震等级预测,具有预测精度较高、考虑因素全面的特点,相比之下SVM模型预测结果略优于FSVM模型。另外,在应用SVM和FSVM进行分类时,如果样本离散性较高,则SVM模型优于FSVM模型;相反,如果样本离散性较低,则FSVM模型优于SVM模型。     

位移反分析的粒子群优化-高斯过程协同优化方法

针对采用随机全局优化技术进行岩土工程位移反分析存在数值计算量大、效率低的问题,将粒子群优化算法与高斯过程机器学习技术相结合,提出了位移反分析的粒子群优化-高斯过程协同优化方法。该方法利用全局寻优性能优异的粒子群优化算法进行寻优的基础上,采用高斯过程机器学习模型不断地总结历史经验,预测包含全局最优解的最有前景区域,通过提高粒子群搜索效率并降低适应度评价次数,进而有效地降低位移反分析过程中的数值计算工作量。多种测试函数的数学验证和工程算例的研究结果表明该方法是可行的,与传统方法相比较,可显著地降低位移反分析的计算耗时。     

基于分频智能反演的储层预测研究——以大庆黑鱼泡南部探区为例

分频反演是近年发展起来的一种先进的地震反演技术。大庆黑鱼泡南部探区存在鼻状构造,是油气富集有利部位。本文基于分频智能反演的储层预测研究,根据地震资料频谱确定分频参数：低频在8 Hz左右,主频35 Hz左右,高频60 Hz左右。根据地震频谱计算得到具有一定带宽的低、中、高频属性。运用支持向量机算法对目标曲线进行分频智能学习训练,学习后曲线相关性达到0.90,满足储层预测的要求。最终,本文将重构曲线与三维地震数据的非线性映射关系引入到二维地震进行分频智能反演,得到二维区阻抗体,对砂岩的预测性较好,为后续的地质建模提供了良好的基础输入数据。