基于TBM破岩数据的岩体条件深度学习表征方法

基于TBM施工数据进行围岩感知对保障TBM施工安全、提高施工效率至关重要,其中TBM掘进参数预测的准确率是检验围岩感知效果的重要依据.为此,以吉林引松工程TBM四标段为研究对象,选取TBM上升段破岩数据为输入特征X1,选择两个施工控制参数（刀盘转速和推进速度）为输入特征X2,构建卷积神经网络机器学习模型,对TBM掘进响应参数Y（刀盘扭矩和总推力）进行预测.按照学习对象的不同,分别构建了只学习稳定段掘进响应行为的点预测模型和同时学习上升段和稳定段掘进响应行为的线预测模型,结果表明：点预测模型无法描述控制参数对掘进响应参数的影响;线预测模型虽然可以描述控制参数对掘进响应参数的影响,但是对稳定段的掘进响应预测数值偏低.考虑到上述局限性的原因是稳定段行为样本数量只占总样本数量的9%,提出了一种通过调节损失函数的方法来提高稳定段行为样本的权重,显著提高了线预测模型的预测精度.改进后的结果表明：在TBM掘进参数预测中,应对整个掘进段的行为进行学习,并提高稳定段行为的权重,以便获得高精度的掘进响应参数预测模型.获得的模型能够为进一步的围岩感知和控制参数优化提供基础.     

利用PLSR-DNN耦合模型预测TBM净掘进速率

科学预测隧道掘进机(TBM)净掘进速率,对于隧道(洞)工程施工方法选择、施工进度安排以及成本估计具有重要意义。鉴于TBM施工过程具有高度非线性、模糊性和复杂性等特征,为提高TBM净掘进速率的预测精度和计算效率,采用偏最小二乘回归(PLSR)提取影响参数主成分,再利用深度神经网络(DNN)进行训练预测,提出了一种基于PLSR-DNN耦合方法的TBM净掘进速率预测模型。基于兰州水源地建设工程输水隧洞双护盾TBM施工实测数据,选择岩石单轴抗压强度、单轴抗拉强度、刀盘推力、刀盘转速、岩体完整性系数和岩石耐磨性指数,共6个影响参数,验证了模型预测的合理性,并对不同预测方法的拟合精度和预测精度进行了对比分析。研究结果表明:(1)偏最小二乘回归可有效克服自变量之间的多重共线性问题,将提取的主成分作为深度神经网络的输入层进行训练,简化了神经网络结构;(2)PLSR-DNN耦合预测模型避免了过拟合与拟合不足问题,具有收敛速度快,求解稳定和拟合精度高等特点;(3)PLSR-DNN耦合预测模型平均相对拟合误差2.96%,平均相对预测误差3.27%,其拟合精度和预测精度均明显高于偏最小二乘回归模型、BP神经网络模型以及支持向量回归(SVR)模型。     

关于用岩体分类预测TBM掘进速率AR的讨论

以Maen、Pieve、Cogolo和Varzo 4条隧道为例,本文分析说明了岩体质量分类系统不能预测TBM净掘进速率PR的原因:由于影响岩体可掘进性的因素与影响岩体质量的因素不一样,且岩体质量分类系统多采用半定性半定量的评分形式描述岩体条件,故岩体的质量分数值与TBM的净掘进速率PR很难有一一对应的关系。因此,岩体质量分类系统和基于此的预测模型不能够用来预测TBM的净掘进速率PR。根据TBM施工隧道岩体分类的目的及TBM施工的特点,提出预测TBM净掘进速率PR的岩体分类系统应与评价岩体稳定的岩体分类系统分开进行。用岩体可掘进性分类系统预测净掘进速率PR,用岩体质量分类系统预测TBM利用率U,从而计算出TBM的掘进速率AR。     

基于随机配点法的地下水污染风险评价

地下水污染事件难以被直接观测,其发生环境通常也具有多种不确定性,风险评价可以量化地下水污染事件的危害性。利用随机配点模型和多项式抽样技术,建立了高效的风险评价模型;考虑了渗透系数、孔隙率和弥散度等多种不确定性因子,探讨了复杂条件下的风险函数的分布特征。研究结果表明:基于随机配点法的风险评价模型避免了多次重复求解对流弥散方程,通过计算成本低廉的随机配点技术得到浓度随机场的拉格朗日多项式,进行多项式抽样获取浓度样本并得到风险函数;与传统的解析算法相比,该方法无需对输入参数和浓度的分布形态做出假设;与传统的蒙特卡罗(Monte Carlo)算法相比,该模型具有明显的效率优势和优越的收敛速度;输入参数的分布类型对风险分布产生显著影响。     

南水北调西线TBM掘进下隧洞围岩损伤研究

为研究隧洞掘进机（tunnel boring machine,TBM）掘进下隧洞围岩的损伤特征,选取南水北调西线工程为背景工程,针对隧洞围岩TBM掘进后的损伤进行了系统的研究与分析。根据板岩在室内试验中表现出来的脆性和屈服后强度特性,提出了以黏聚力弱化-摩擦角强化（cohesion weakening and frictional strengthening,CWFS）脆性模型为基础考虑岩体屈服后损伤的本构模型,通过VC++编译dll动态链接库,并将其植入FLAC3D程序中。应用神经网络和遗传方法等优化算法对本构模型中的参数进行了反演分析,并以室内三轴压缩试验为依据,对该本构模型进行了验证。结果表明,提出的考虑损伤的CWFS模型可以较好地反映板岩的弹性特征、屈服后的强化特征、脆性破坏特征和破坏后残余强度特征。以该模型为基础,模拟了南水北调西线隧洞在TBM掘进条件下隧洞围岩的变形特性、应力发展规律及其损伤演化特征,并与基于Mohr-Coulomb本构关系的计算结果进行了对比分析。提出的考虑损伤的CWFS模型能更好地反映隧洞掘进后围岩损伤特征,即损伤沿着弱势区域发展的规律     

二维随机场下边坡稳定性的径向基函数神经网络分析法

岩土参数的随机性会直接影响边坡稳定性评价结果的精度。首先,依据边坡参数的常用分布特征,利用拉丁超立方抽样法生成若干组边坡土性参数和几何参数的随机样本,用有限元强度折减法求解各组样本对应的边坡安全系数。再考虑土性参数的空间变异性,在二维随机场模型下将蒙特卡罗模拟和有限元强度折减法相结合求解各组样本对应的边坡失效概率。然后,利用样本数据及其安全系数和失效概率对径向基函数（RBF）神经网络进行训练和测试,从而建立边坡安全系数和失效概率的预测模型。算例表明,二维随机场模型能相对精确地考虑参数的空间变异性;在此基础上建立的神经网络模型对边坡的安全系数和失效概率具有较高的预测精度,且能极大地节省边坡稳定性分析的时间。     

隧道掘进爆破地震峰值神经网络预报研究

结合某抽水蓄能电站尾水隧道掘进爆破地面振动监测结果,建立了基于BP神经网络隧道掘进爆破振动速度峰值的预报模型。并对铁路运输安全进行预报,与现场实际较好地吻合。将神经网络模型预报的结果与传统方法(经验公式法)预报的结果相比,前者的预报结果有明显的改善。对于指导隧道掘进爆破设计,优化爆破参数,确保安全具有重要的意义。     

岩石矿物细胞元随机性参数赋值方法研究

数值计算中利用随机分布对各单元赋予不同的物理力学参数来考虑岩石的非均质性,赋值过程是纯随机的,并没有考虑岩石矿物组成的结构特征。故提出一种新的岩石非均质参数赋值方法--岩石矿物细胞元随机性参数赋值方法,基于岩石矿物种类及其含量定义细胞元类别判定区间,利用Monte Carlo方法对各个细胞元进行矿物类别判定,并进行相应的参数赋值,通过各矿物细胞元的随机混合体来描述岩石的非均质性。该方法既考虑了组成岩石的矿物种类及其含量（结构特征）,又考虑了组成矿物在岩石中的随机分布特征。针对矿物分布的随机特征,利用两矿物细胞元混合模型和三矿物细胞元混合模型进行数值试验,研究了矿物细胞元随机分布特征对岩石宏观力学参数的影响。研究结果表明,岩石矿物细胞元随机性参数赋值方法具有结构性和随机性的双重特性,其随机性不依赖于随机参数,岩石宏观力学参数受细胞元随机性特征的影响很小。     

采空区移动变形的径向基函数神经网络概率积分法(RBF)反演

借助部分矿区地表移动实测数据,选用径向基函数(RBF)神经网络对概率积分法的计算参数进行反演,采用K层交叉验证对模型精度进行优化,并与BP神经网络模型和SVM模型预测结果进行对比,发现RBF模型精度均优于BP模型和SVM模型,且其稳定性较好。可为概率积分法预测评估采空区地表移动变形范围提供一种可靠的方法。     

BP 模型在南海神狐海域天然气水合物储量参数预测中的应用

在用测井数据预测储量参数方法的基础上,采用BP 神经网络法预测天然气水合物储量参数( 孔隙度、饱和度) 。选取一口有实测值的井,将其测井数据作为样本数据,建立网络模型,由其他井的测井数据输入此模型得到储量参数预测结果。经过实践检验此模型得出的结果比经验公式法更精确。     

Development and validation of a 3D numerical model for TBM–EPB mechanised excavations

A 3D numerical model for mechanised excavations is presented, which is capable of simulating the overall process of excavation and construction of a tunnel when a TBM EPB (Tunnel Boring Machine–Earth Pressure Balance) is used.The main construction aspects of a mechanised excavation are modelled. Their influence on calculated ground displacements are investigated by means of a series of parametric analyses.With the aim of testing the performance of the proposed 3D numerical model, a series of 25 Class C predictions has been carried out. Case Histories related to the construction of the 1995–2003 Madrid Metro Extension Project were considered for this purpose.As a general rule, the results obtained with the Modified Cam–Clay model closely fit in situ measurements. When the Linear-Elastic or the Mohr–Coulomb models are used, it is not as easy to summarise the results obtained, as higher fluctuations are observed around in situ measured data. A good agreement is also shown when the distribution of horizontal displacements along depth is considered.For some sections, the mechanised excavation model is not capable of reproducing the high values of the surface settlements measured in situ. A closer look at the results shows that mixed face conditions are found for these cases, with the TBM excavating through layered soil formations having sharply different mechanical behaviour.     

隧洞变形引起的TBM施工事故综合风险分析

国内采用岩石隧道掘进机（简称TBM）施工的隧洞呈现深埋、超长趋势,变形引起的TBM卡机事故屡见不鲜,为了减少这些事故,可以预先对变形引起的TBM风险事故进行综合评价,因此,在单元风险研究的基础上对综合风险计算及评价准则进行研究。采用概率理论推导了变形引起的TBM风险事故的综合风险概率的计算模型,根据推导的模型可以近似计算出5类后果等级事故发生的概率;采用TBM的卡机时间与纯掘进时间的比值作为分级指标,结合以往TBM施工的统计数据划分工期损失后果等级,再结合风险的概率分级,建立变形综合风险评价准则;采用研究的理论,对工程选取段TBM施工卡机风险进行了两种工况下的计算分析。计算结果表明,施工时若不采取任何措施TBM施工风险比较大,但采用合理的措施之后风险可控制在可接受的范围内。     

TBM Performance Analysis in Pyroclastic Rocks: A Case History of Karaj Water Conveyance Tunnel

Karaj Water Conveyance Tunnel (KWCT) is 30-km long and has been designed for transferring 16 m³/s of water from Amir-Kabir dam to northwest of Tehran. Lot No. 1 of this long tunnel, with a length of 16 km, is under construction with a double shield TBM and currently about 8.7 km of the tunnel has been excavated/lined. This paper will offer an overview of the project, concentrating on the TBM operation and will review the results of field performance of the machine. In addition to analysis of the available data including geological and geotechnical information and machine operational parameters, actual penetration and advance rates will be compared to the estimated machine performance using prediction models, such as CSM, NTNU and Q_TBM. Also, results of analysis to correlate TBM performance parameters to rock mass characteristics will be discussed. This involves statistical analysis of the available data to develop new empirical methods. The preliminary results of this study revealed that the available prediction models need some corrections or modifications to produce a more accurate prediction in geological conditions of this particular project.     

Applying Optimized Support Vector Regression Models for Prediction of Tunnel Boring Machine Performance

One of the main factors in the effective application of a tunnel boring machine (TBM) is the ability to accurately estimate the machine performance in order to determine the project costs and schedule. Predicting the TBM performance is a nonlinear and multivariable complex problem. The aim of this study is to predict the performance of TBM using the hybrid of support vector regression (SVR) and the differential evolution algorithm (DE), artificial bee colony algorithm (ABC), and gravitational search algorithm (GSA). The DE, ABC and GSA are combined with the SVR for determining the optimal value of its user defined parameters. The optimization implementation by the DE, ABC and GSA significantly improves the generalization ability of the SVR. The uniaxial compressive strength (UCS), average distance between planes of weakness (DPW), the angle between tunnel axis and the planes of weakness (α), and intact rock brittleness (BI) were considered as the input parameters, while the rate of penetration was the output parameter. The prediction models were applied to the available data given in the literature, and their performance was assessed based on statistical criteria. The results clearly show the superiority of DE when integrated with SVR for optimizing values of its parameters. In addition, the suggested model was compared with the methods previously presented for predicting the TBM penetration rate. The comparative results revealed that the hybrid of DE and SVR yields a robust model which outperforms other models in terms of the higher correlation coefficient and lower mean squared error.     

Prediction of time to slope failure: a general framework

The prediction of time to slope failure (TSF) is a goal of major importance for both landslide researchers and practitioners. A reasonably accurate prediction of TSF allows human losses to be avoided, damages to property to be reduced and adequate countermeasures to be designed. A pure “phenomenological” approach based on the observation and interpretation of the monitored data is generally employed in TSF prediction. Such an approach infers TSF mainly from the ground surface displacements using regression techniques based on empirical functions. These functions neglect the rheological soil parameters in order to reduce the prediction uncertainties. This paper presents an overlook of the methods associated with this approach and proposes a unique expression encompassing most of the previously proposed equations for TSF prediction, thus offering a general framework useful for comparisons between different methods. The methods discussed in this paper provide an effective tool, and sometimes the only tool, for TSF prediction. The fundamental problem is always one of data quality. A full confidence in all assumptions and parameters used in the prediction model is rarely, if ever, achieved. Therefore, TSF prediction models should be applied with care and the results interpreted with caution. Documented case studies represent the most useful source of information to calibrate the TSF prediction models.     

基于随机理论的地下水环境风险评价

针对确定性模型难以描述含水层非均质空间分布的问题,提出基于随机理论的地下水环境风险评价方法。以矩形场地地下水污染风险评价为例,采用蒙特卡罗法生成大量渗透系数随机场,模拟含水层参数各种可能的非均质空间分布,在此基础上建立场地地下水流模型与溶质运移模型,分别计算污染物在地下水中的迁移转化情况。统计大量随机模拟中污染事故发生的频率,当模拟次数足够多时,污染频率收敛于污染概率,污染风险即通过污染概率体现出来。该方法将模型参数设为满足一定分布特征的随机变量,避免了确定性方法得出的武断的评价结果,可为工厂的选址、水源地的选址等工作提供科学指导。     

全断面隧道掘进机护盾受力监测及卡机预警

全断面隧道掘进机(简称TBM)在穿越深部软弱地层时围岩收敛变形较大,围岩容易挤压护盾,导致TBM卡机,影响TBM正常掘进。通过分析TBM卡机灾害孕育过程,得到了预测TBM卡机的重要条件：一是围岩变形量大于预留的空间,二是额定推力不能克服摩擦阻力。为了监测实际工程TBM卡机状态,提出了一种监测护盾变形的方案以及护盾受力的计算方法,可通过监测得到的变形估算护盾的受力,进而计算出护盾受到的摩擦阻力,得到TBM卡机的状态。根据TBM受到的摩擦阻力、TBM正常掘进时所需推力和TBM额定推力之间的关系,将TBM卡机状态分为4个等级,即无卡机、轻微卡机、卡机和严重卡机,并提出了对应的处理措施。结合TBM卡机条件以及护盾受力监测方案,提出了TBM卡机灾害预警流程。在兰州水源地输水隧洞工程中应用了该监测方案和卡机灾害预警流程,应用结果表明,预测的卡机状态与TBM实际状态基本一致,说明该方法具有一定的可靠性,对指导TBM隧道施工具有重要意义。     

Universal model forms for predicting the shear modulus and material damping of soils

Numerous model forms have been used to predict the dynamic behaviours of soils, mainly the shear modulus and material damping. These models are used to represent normalised shear modulus reduction and material damping curves. The majority of the existing models are empirical and were proposed for specific soil types, strain ranges, etc. Some are limited to the data used in fitting, but fail to provide a good fit to other sets of data. The available model functions are not universal, which means that a model developed for one soil type may not be applicable for use with other soil types. In this paper, two universal mathematical models were proposed to predict the normalised shear modulus reduction and material damping curves. The mathematical model forms are sufficiently flexible to be widely used with different soil types. The models were evaluated using verification data of eight different soil types, i.e., data that were not used to fit the model parameters. The results indicate that the proposed model forms can accurately model the dynamic soil properties within the typical earthquake range. The advantage of developing the two universal models are two-fold, first the forms are applicable for sand, clay, and fibre-soil composite and second with further testing of soils, the model coefficients could ultimately be used to further understand the physical processes in soil behaviour, especially damping.     

考虑邻近土洞影响的盾构掘进速度控制

盾构机穿越邻近土洞区域是在岩溶区修建的地铁隧道常常遇到的问题,为降低盾构掘进对土洞的扰动,以防止盾构突陷、偏离轴线或发生地面塌陷,需对穿越时盾构的掘进参数进行优化控制。针对当前主要根据经验选取掘进参数的不足,根据弹性力学Mindlin解建立了盾构穿越邻近土洞的力学模型,在此基础上,以掘进速度为控制变量,以掘进引起土洞顶部能量密度变化为指标函数,提出了穿越时的掘进速度最优控制问题,并利用梯度法进行了数值求解,最后将该方法运用于广州地铁九号线花-马区间盾构穿越邻近土洞问题的分析中。结果表明:该模型能有效反映盾构掘进时正面推力对土洞顶部的扰动;在该模型上建立的掘进速度最优控制问题,能对掘进速度的控制策略进行优化,优化结果与工程经验相符,具有一定工程应用价值。     

Prediction of blasting-induced fragmentation in Meydook copper mine using empirical,statistical, and mutual information models

One of the most important aims of blasting in open pit mines is to reach desirable size of fragmentation. Prediction of fragmentation has great importance in an attempt to prevent economic drawbacks. In this study, blasting data from Meydook mine were used to study the effect of different parameters on fragmentation; 30 blast cycles performed in Meydook mine were selected to predict fragmentation where six more blast cycles are used to validate the results of developed models. In this research, mutual information (MI) method was employed to predict fragmentation. Ten parameters were considered as primary ones in the model. For the sake of comparison, Kuz-Ram empirical model and statistical modeling were also used. Coefficient of determination (R ²), root mean square error (RMSE), and mean absolute error (MAE) were then used to compare the models. Results show that MI model with values of R ², RMSE, and MAE equals 0.81, 10.71, and 9.02, respectively, is found to have more accuracy with better performance comparing to Kuz-Ram and statistical models.