Prediction of rockhead using a hybrid N-XGBoost machine learning framework

The spatial information of rockhead is crucial for the design and construction of tunneling or underground excavation. Although the conventional site investigation methods (i.e. borehole drilling) could provide local engineering geological information, the accurate prediction of the rockhead position with limited borehole data is still challenging due to its spatial variation and great uncertainties involved. With the development of computer science, machine learning (ML) has been proved to be a promising way to avoid subjective judgments by human beings and to establish complex relationships with mega data automatically. However, few studies have been reported on the adoption of ML models for the prediction of the rockhead position. In this paper, we proposed a robust probabilistic ML model for predicting the rockhead distribution using the spatial geographic information. The framework of the natural gradient boosting (NGBoost) algorithm combined with the extreme gradient boosting (XGBoost) is used as the basic learner. The XGBoost model was also compared with some other ML models such as the gradient boosting regression tree (GBRT), the light gradient boosting machine (LightGBM), the multivariate linear regression (MLR), the artificial neural network (ANN), and the support vector machine (SVM). The results demonstrate that the XGBoost algorithm, the core algorithm of the probabilistic N-XGBoost model, outperformed the other conventional ML models with a coefficient of determination (R²) of 0.89 and a root mean squared error (RMSE) of 5.8 m for the prediction of rockhead position based on limited borehole data. The probabilistic N-XGBoost model not only achieved a higher prediction accuracy, but also provided a predictive estimation of the uncertainty. Thus, the proposed N-XGBoost probabilistic model has the potential to be used as a reliable and effective ML algorithm for the prediction of rockhead position in rock and geotechnical engineering.     

Effectiveness of predicting tunneling-induced ground settlements using machine learning methods with small datasets

Prediction of tunneling-induced ground settlements is an essential task, particularly for tunneling in urban settings. Ground settlements should be limited within a tolerable threshold to avoid damages to aboveground structures. Machine learning(ML) methods are becoming popular in many fields, including tunneling and underground excavations, as a powerful learning and predicting technique. However, the available datasets collected from a tunneling project are usually small from the perspective o...     

Poroelastic behaviour of physical properties in Rotliegend sandstones under uniaxial strain

Laboratory experiments simulating reservoir depletion and reinjection of fluids have been conducted on Rotliegend reservoir sandstones. Pore pressure dependence of petrophysical properties have been measured under uniaxial strain boundary conditions, i.e., zero lateral strain at constant overburden pressure. Permeability, formation resistivity factor, and compressional as well as shear wave velocities were recorded simultaneously and continuously during deformation in direction of maximum principal stress. The results show that the stress development is specific for different sandstones depending on the efficiency of pore pressure. The stress anisotropy increases with decreasing pore pressure, leading to stress-induced anisotropy of pore structure with preferred closure of horizontally oriented pore space. Permeability decline of initially lower permeability sandstones indicate the opening of axially oriented pores at lower pore pressure.     

Structural performance assessment of GFRP elastic gridshells by machine learning interpretability methods

The prediction of structural performance plays a significant role in damage assessment of glass fiber reinforcement polymer (GFRP) elastic gridshell structures. Machine learning (ML) approaches are implemented in this study, to predict maximum stress and displacement of GFRP elastic gridshell structures. Several ML algorithms, including linear regression (LR), ridge regression (RR), support vector regression (SVR), K-nearest neighbors (KNN), decision tree (DT), random forest (RF), adaptive boosting (AdaBoost), extreme gradient boosting (XGBoost), category boosting (CatBoost), and light gradient boosting machine (LightGBM), are implemented in this study. Output features of structural performance considered in this study are the maximum stress as f₁(x) and the maximum displacement to self-weight ratio as f₂(x). A comparative study is conducted and the Catboost model presents the highest prediction accuracy. Finally, interpretable ML approaches, including shapely additive explanations (SHAP), partial dependence plot (PDP), and accumulated local effects (ALE), are applied to explain the predictions. SHAP is employed to describe the importance of each variable to structural performance both locally and globally. The results of sensitivity analysis (SA), feature importance of the CatBoost model and SHAP approach indicate the same parameters as the most significant variables for f₁(x) and f₂(x).     

Prediction of embankment behavior of regulating reservoir with foundation improved by vacuum consolidation method

The present paper addresses the numerical prediction of the behavior of a ground and a reservoir dyke with a retaining wall at the site of a regulating reservoir whose soft soil foundation is improved by using both the usual embankment preloading and vacuum consolidation. To evaluate the settlement, the lateral deformation, and the dissipation of pore pressure during vacuum preloading, a numerical analysis using an elasto-plastic FEM for soil–water coupled problems, incorporating the SYS Cam-clay model, is carried out in two dimensions. However, a change in the soil parameters during the vacuum preloading leads to a less accurate computation. To account for the uncertainties in the input parameters of the constitutive model for the improved ground, an inverse analysis approach is adopted. The particle filter is used to identify the compression index of the clay layers and the coefficient of permeability of the organic soil layer based on the measured settlement at the bottom of the preloading embankment during the vacuum consolidation. The reservoir dyke with a retaining wall is constructed on an improved foundation after removing the preloading embankment, and an attempt is made to predict its performance after construction by an elasto-plastic FEM for soil–water coupled problems using the identified parameters.     

Comparison of machine learning methods for ground settlement prediction with different tunneling datasets

This study integrates different machine learning (ML) methods and 5-fold cross-validation (CV) method to estimate the ground maximal surface settlement (MSS) induced by tunneling. We further investigate the applicability of artificial intelligent (AI) based prediction through a comparative study of two tunnelling datasets with different sizes and features. Four different ML approaches, including support vector machine (SVM), random forest (RF), back-propagation neural network (BPNN), and deep neural network (DNN), are utilized. Two techniques, i.e. particle swarm optimization (PSO) and grid search (GS) methods, are adopted for hyperparameter optimization. To assess the reliability and efficiency of the predictions, three performance evaluation indicators, including the mean absolute error (MAE), root mean square error (RMSE), and Pearson correlation coefficient (R), are calculated. Our results indicate that proposed models can accurately and efficiently predict the settlement, while the RF model outperforms the other three methods on both datasets. The difference in model performance on two datasets (Datasets A and B) reveals the importance of data quality and quantity. Sensitivity analysis indicates that Dataset A is more significantly affected by geological conditions, while geometric characteristics play a more dominant role on Dataset B.     

Investigation of feature contribution to shield tunneling-induced settlement using Shapley additive explanations method

Accurate prediction of shield tunneling-induced settlement is a complex problem that requires consideration of many influential parameters. Recent studies reveal that machine learning(ML) algorithms can predict the settlement caused by tunneling. However, well-performing ML models are usually less interpretable. Irrelevant input features decrease the performance and interpretability of an ML model. Nonetheless, feature selection, a critical step in the ML pipeline, is usually ignored in most stu...     

A generic framework for geotechnical subsurface modeling with machine learning

This study introduces a generic framework for geotechnical subsurface modeling, which accounts for spatial autocorrelation with local mapping machine learning(ML) methods. Instead of using XY coordinate fields directly as model input, a series of autocorrelated geotechnical distance fields(GDFs) is designed to enable the ML models to infer the spatial relationship between the sampled locations and unknown locations. The whole framework using GDF with ML methods is named GDF-ML. This framework is...     

Evaluation and prediction of slope stability using machine learning approaches

In this paper, the machine learning (ML) model is built for slope stability evaluation and meets the high precision and rapidity requirements in slope engineering. Different ML methods for the factor of safety (FOS) prediction are studied and compared hoping to make the best use of the large variety of existing statistical and ML regression methods collected. The data set of this study includes six characteristics, namely unit weight, cohesion, internal friction angle, slope angle, slope height, and pore water pressure ratio. The whole ML model is primarily divided into data preprocessing, outlier processing, and model evaluation. In the data preprocessing, the duplicated data are first removed, then the outliers are filtered by the LocalOutlierFactor method and finally, the data are standardized. 11 ML methods are evaluated for their ability to learn the FOS based on different input parameter combinations. By analyzing the evaluation indicators R ², MAE, and MSE of these methods, SVM, GBR, and Bagging are considered to be the best regression methods. The performance and reliability of the nonlinear regression method are slightly better than that of the linear regression method. Also, the SVM-poly method is used to analyze the susceptibility of slope parameters.     

Effect of soil properties for prediction of energy consumption in land levelling irrigation

ABSTRACT

Energy conservation is regarded as one of the most important factors in the agricultural sector due to its relation to pollution which is a result of fossil fuel (particularly gasoline) usage. The objective of this research was to develop three methods including Artificial Neural Network (ANN), regression and Adaptive Neural-Fuzzy Inference System (ANFIS) to predict the effect of soil properties on the environmental indicators in land levelling and to analyse the sensitivity of these parameters. The acquired data were used to develop accurate models for fuel energy (FE), total machinery cost (TMC) and total machinery energy (TME). The results showed that four parameters of tillage depth, forward speed, cone index and cut/fill volume had significant effects on energy consumption. ANFIS and ANN had a satisfactory performance in predicting the aforementioned parameters in the various field conditions. The ANN had the most capability in FE prediction according to the least RMSE and the highest coefficient of determination (R²) values 0.0206 and 0.9983, respectively. The ANFIS model had the most capability in the prediction of the environmental and energy parameters with the least RMSE and the highest R2 for TMC, 0.0287 and 0.9966, and for TME, 0.0157 and 0.9990, respectively.     

Failure criterion effect on solid production prediction and selection of completion solution

Production of fines together with reservoir fluid is called solid production. It varies from a few grams or less per ton of reservoir fluid posing only minor problems, to catastrophic amount possibly leading to erosion and complete filling of the borehole. This paper assesses solid production potential in a carbonate gas reservoir located in the south of Iran. Petrophysical logs obtained from the vertical well were employed to construct mechanical earth model. Then, two failure criteria, i.e. Mohre Coulomb and Mogi-Coulomb,were used to investigate the potential of solid production of the well in the initial and depleted conditions of the reservoir. Using these two criteria, we estimated critical collapse pressure and compared them to the reservoir pressure. Solid production occurs if collapse pressure is greater than pore pressure. Results indicate that the two failure criteria show different estimations of solid production potential of the studied reservoir. Mohre Coulomb failure criterion estimated solid production in both initial and depleted conditions, where Mogi-Coulomb criterion predicted no solid production in the initial condition of reservoir. Based on Mogi-Coulomb criterion, the well may not require completion solutions like perforated liner, until at least 60% of reservoir pressure was depleted which leads to decrease in operation cost and time.     

Characterization of 3D pore nanostructure and stress-dependent permeability of organic-rich shales in northern Guizhou Depression,China

The three-dimensional (3D) pore structures and permeability of shale are critical for forecasting gas production capacity and guiding pressure differential control in practical reservoir extraction. However, few investigations have analyzed the effects of microscopic organic matter (OM) morphology and 3D pore nanostructures on the stress sensitivity, which are precisely the most unique and controlling factors of reservoir quality in shales. In this study, ultra-high nanoscale-resolution imaging experiments, i.e. focused ion beam-scanning electron microscopy (FIB-SEMs), were conducted on two organic-rich shale samples from Longmaxi and Wufeng Formations in northern Guizhou Depression, China. Pore morphology, porosity of 3D pore nanostructures, pore size distribution, and connectivity of the six selected regions of interest (including clump-shaped OMs, interstitial OMs, framboidal pyrite, and microfractures) were qualitatively and quantitatively characterized. Pulse decay permeability (PDP) measurement was used to investigate the variation patterns of stress-dependent permeability and stress sensitivity of shales under different confining pressures and pore pressures, and the results were then used to calculate the Biot coefficients for the two shale formations. The results showed that the samples have high OM porosity and 85% of the OM pores have the radius of less than 40 nm. The OM morphology and pore structure characteristics of the Longmaxi and Wufeng Formations were distinctly different. In particular, the OM in the Wufeng Formation samples developed some OM pores with radius larger than 500 nm, which significantly improved the connectivity. The macroscopic permeability strongly depends on the permeability of OM pores. The stress sensitivity of permeability of Wufeng Formation was significantly lower than that of Longmaxi Formation, due to the differences in OM morphology and pore structures. The Biot coefficients of 0.729 and 0.697 were obtained for the Longmaxi and Wufeng Formations, respectively.     

灰色GM(1,1)模型在真空预压孔隙水压力分析中应用

真空预压法是加固软土地基行之有效的方法,结合工程实例,分析了灰色GM(1,1)模型在真空预压孔隙水压力预测中的可行性,并根据预测结果计算了孔隙水压力固结度,结果表明灰色GM(1,1)模型预测结果准确可信,分析评价结果对工程施工具有重要的指导意义。     

基于承压含水层水力响应的溪洛渡水电工程区谷幅收缩变形预测研究

溪洛渡高拱坝位于一个向斜盆地内,区域水文地质的一个典型特征是盆地内下覆有一个完整延伸的阳新灰岩承压含水层。水库蓄水以来上下游库岸边坡发生了明显的谷幅收缩变形,主要是由于水库蓄水引起承压含水层孔隙水压力增加,从而引起灰岩地层因有效应力减小而产生膨胀,以及相对隔水层底板扬压力增加所引起的。基于承压含水层水力响应规律,建立了谷幅收缩变形反演及预测的解析模型,分别预测了在变形速率小于0.01,0.001mm/d两个收敛准则下的谷幅收缩变形收敛时间以及收敛值。所建立的反演预测模型能够较好地重现谷幅收缩变形历时曲线,反映出谷幅收缩变形与库水位变化具有高度的相关性。模型预测结果显示,溪洛渡水电工程的谷幅收缩变形已趋于收敛。     

Forecasting construction industry demand,price and productivity in Singapore: the BoxJenkins approach

In academic research, the traditional Box-Jenkins approach is widely acknowledged as a benchmark technique for univariate methods because of its structured modelling basis and acceptable forecasting performance. This study examines the versatility of this approach by applying it to analyse and forecast three distinct variables of the construction industry, namely, tender price, construction demand and productivity, based on case studies of Singapore. In order to assess the adequacy of the Box-Jenkins approach to construction industry forecasting, the models derived are evaluated on their predictive accuracy based on out-of-sample forecasts. Two measures of accuracy are adopted, the root mean-square-error (RMSE) and the mean absolute percentage error (MAPE). The conclusive findings of the study include: (1) the prediction RMSE of all three models is consistently smaller than the model's standard error, implying the models' good predictive performance; (2) the prediction MAPE of all three models consistently falls within the general acceptable limit of 10%; and (3) among the three models, the most accurate is the demand model which has the lowest MAPE, followed by the price model and the productivity model.     

Model-scale tests to examine water pressures acting on potentially buoyant underground structures in clay strata

Throughout the service life, underground structures are subjected to transient and sustained hydrostatic pressures. The reservoir impoundment results in an increase in water level, as well as hydraulic gradient, which can endanger the uplift performance of infrastructure. In uplift design, a reduction factor is often suggested for buoyant force acting on underground structures in clays due to the time lag effect. However, the mechanism of pore pressure generation in clays is not fully understood. This investigation presents a novel U-shaped test chamber to assess the pore pressure generation with time in the horizontal branch subjected to an increase in reservoir level in the left vertical branch. A mathematical model is developed to explain the time lag effect of pore pressure generation. The test program also involves the evaluation of uplift pressure acting on foundation model in the right vertical branch due to adjacent reservoir impoundment. It is found that the time lag effect of pore pressure generation in clays can be observed irrespective of hydraulic gradient, but a higher hydraulic gradient can lead to a faster response in pore pressure sensors. A reduction factor of 0.84–0.87 should be considered to reduce the conservatism of uplift design.     

鄂尔多斯盆地延长组致密砂岩储层微观孔隙结构特征

孔隙结构是研究储层物性的主要内容.该文在调研前人研究的基础上,分析总结了鄂尔多斯盆地延长组致密砂岩储层微观孔隙结构特征.结果 表明:前人对鄂尔多斯盆地延长组储层微观孔隙结构表征参数选取不统一,一般选取能够代表研究区微观孔喉结构及渗流特征的优选参数,建立致密储层的分类评价标准.对致密油的孔隙结构表征方法主要有3种:数据统...     

Dynamic thermal building analysis with CFD – modelling radiation

In an attempt to reduce the high computational effort required for dynamic thermal simulation of buildings using computational fluid dynamics (CFD) the authors have recently developed an adaptive freeze-flow method (i.e. freezing of flow equations over variable time periods). This article documents the work that has been carried out to predict the surface heat transfer in dynamic thermal building processes using CFD with particular focus on radiation. The Monte Carlo (MC) and discrete transfer (DT) radiation models were investigated and results compared with analytical solutions. The DT model has shown good performance whereas an unrealistic radiation distribution on the surfaces was observed when using the MC model. A further investigation of the DT model for the cooling of a solid wall has shown that the adaptive freeze-flow method is an efficient and accurate means of conducting dynamic thermal CFD simulations which involve radiation. Finally, application of the technique to a more realistic space comprising an uneven distribution of solar gain showed very good results when compared with a zonal dynamic thermal simulation program.     

两种循环荷载共同作用下地基一维固结

一维固结理论是本文的一个特例。根据不同土层深度以及地下水位的变化对孔压变化的影响进行固结性状分析,得到有关的固结曲线,结果表明:两种循环荷载作用下,固结开始时固结速度反而慢,但是随着固结的深入,其固结速率加快,孔压值达到一定值后,与单种循环荷载作用趋于一致;两种循环荷载下孔压值随时间先增大后减小,与实验结果先一致;孔压呈振荡变化并且最终趋向于某一定值后基本保持不变。     

Geomechanical response of overburden caused by CO_2 injection into a depleted oil reservoir

This study investigates the hydro-mechanical aspects of carbon dioxide(CO_2) injection into a depleted oil reservoir through the use of coupled multiphase fluid flow and geomechanical modeling.Both singlephase and multiphase fluid flow analyses coupled with geomechanics were carried out at the West Pearl Queen depleted oil reservoir site,and modeling results were compared with available measured data.The site geology and the material properties determined on the basis of available geophysical data were used in the analyses.Modeling results from the coupled multiphase fluid flow and geomechanical analyses show that computed fluid pressures match well with available measured data.The hydromechanical properties of the reservoir have a significant influence on computed fluid pressures and surface deformations.Hence,an accurate geologic characterization of the sequestration site and determination of engineering properties are important issues for the reliability of model predictions.The computed fluid pressure response is also significantly influenced by the relative permeability curves used in multiphase fluid flow models.While the multiphase fluid flow models provide more accurate fluid pressure response,single-phase fluid flow models can be used to obtain approximate solutions.The ground surface deformations obtained from single-phase fluid flow models coupled with geomechanics are slightly lower than those predicted by multiphase fluid flow models coupled with geomechanics.However,the advantage of a single-phase model is the simplicity.Limited field monitoring of subsurface fluid pressure and ground surface deformations during fluid injection can be used in calibrating coupled fluid flow and geomechanical models.The calibrated models can be used for investigating the performance of large-scale CO_2 storage in depleted oil reservoirs.