Characterizing multivariate parameters is crucial for uncertainty analysis in geological engineering. However, the commonly used multivariate distribution function-based methods are increasingly challenged by issues of diminished accuracy and efficiency, as well as the inclusion of subjective processes, due to the growing complexity of data structures resulting from advanced data acquisition techniques. To this end, the prevailing generative machine learning algorithms are tried to learn the joint distributions of multivariate geotechnical parameters, and a deep neural network, CasMDN, is proposed based on the mixture density network and a cascade mechanism. The principle, structure, and training method of CasMDN is first introduced. Then two applications of the model, including stochastic simulation and probability calculation, are presented. In the first experiment, a group of geometric parameters of rock joints are simulated by a variational autoencoder, a generative adversarial network, a Gaussian mixture model, and our model, respectively. The result shows a significant advantage of our method over other generative models in learning the joint distribution of orientation, trace length, and aperture. Then, another comparison between our method and the recently proposed copula-based approach is carried out with a set of five-dimensional soil parameters, which is an open-source dataset collected from previous literature. It is proved that the CasMDN model can achieve the best performances on both fitting and probability calculating. This research demonstrates that using intelligent generative models to characterize complex geotechnical parameters can help explore the deep laws of geological uncertainty by improving accuracy and reducing subjectivity. Furthermore, the proposed CasMDN model is considered to have a broad application scope beyond the provided cases, for instance, uncertainty modelling and risk assessment of geotechnical engineering.

Basic geological phenomena in a hydraulic cavern can be a reflection of the occurrence of geological disaster. An efficient classification and identification of such information are crucial to understanding the distribution of structural planes in the cavern and the properties of its surrounding rocks to guide its further exploration. This task usually relies on manual work, but such traditional methods are time-consuming and labor-intensive at a low level of automated analysis. In this work, we use a variety of deep learning models and machine learning models to analyze the images of basic geological observations of hydraulic cavern foundation; and the cavern phenomena are classified through applying different deep learning models and the methods of Softmax classifier, random forest, and support vector machine. By comparing and selecting models with better performance for coupling, we developed a satisfactory image recognition model for the cavern geological phenomena, achieving automatic identification and analysis of the caverns, so that the workload of geological engineers is reduced significantly.

In the eastern plain area of the Haihe River basin of Tianjin, the ground subsidence is serious and the overburden layer is deep. To improve the construction of the watershed elevation system under this terrain condition, four kilometer-level bedrock markers were built for Tuanbowa Reservoir, Qingyun hydrology station, Xianxian Hydrology Station,and Xinjiafang hydrology station, and complete the bedrock the leveling network initial measurements and key hydraulic structure level testing work. According to the investigation, the depth of these four bedrock markers will be ranked in the world’s top four for a considerable period. The construction of kilometer-level bedrock is extremely difficult and requires high technology, which is rare at home and abroad. This paper discusses in detail site selection, construction technology,hole slope control, and rock mass stability, which can provide valuable experience for the subsequent construction.

Aiming at the problems of defects suchas corrosion and fracture of prestressed anchor cables under long-term operation, a theoretical method to determine the defect location of anchor cables is proposed based on the strain monitoring technology and the numerical simulation tests on the middle, left and right defective anchor cables. The experimental results show that there is a relationship between the defect location and the strain, and the results and trends of the numerical simulation and physical tests are basically the same. The relationship between the cable strain and the defect location is established through 15 groups of numerical simulation tests on different defect locations. It provides a feasible way to determine the defect location of cables and also a technical support for long-term safe operation of prestressed anchorage engineering.

The integrity of rock mass is an important parameter in evaluating the quality grade of hydraulic rock mass. The traditional methods often adopt a single index that cannot fully reflect the comprehensive influences of geological conditions such as structural plane, groundwater, and unloading on the evaluation results. An intelligent evaluation method for the integrity of hydraulic rock mass is proposed by coupling with multi-source survey information. Firstly, the synthetic minority oversampling (SMOTE) algorithm is used to balance the survey information data to improve the data set structure. Then the random forest algorithm is used to predict the original rock mass integrity data and the pre-processed data, respectively. Based on the data of actual projects, the validity and applicability are verified, and the predicted results are discussed and analyzed according to different factors affecting the integrity of rock mass. The results show that the proposed method can effectively improve the evaluation accuracy of the integrity of a few rock samples by balancing the data sets. By coupling and mining the deep information of different integrity indexes, the intelligent evaluation of rock mass integrity can be realized, which provides a new method for further assisting the evaluation of rock mass quality.

Geological investigation is a basic work in hydraulic project construction. The traditional geological survey records geological information on paper. However, this leads to a poor intuitiveness of the two-dimensional map and even a loss of its spatial features, and the field analysis process has to rely on subjective experiences heavily; disconnection of office work from field work makes it difficult to ensure data recording and analysis efficiency or data reliability. This paper presents an integrated method of 3D real scene recording and office-field analysis for hydraulic engineering geology. This system is based on the 3D real scene technology, Geographic Information System（GIS）, machine learning algorithm, and geological modeling, focusing on the development of multi-source data fusion, multiplatform collaboration and 2D and 3D linkage, geological intelligence recording, and fast spatial interpretation. We develop two sets of platforms suitable for desktop and mobile terminals, including three main modules: 1) A base map module constructs a fused 3D reality using multi-source data and achieves multi-platform collaboration as well as 2D and 3D linkage of data, thereby providing reference scenes for office work and field work; 2) A recording module adopts a multi-platform collaboration and intelligent recognition method to implement the recording of geological points, geological boundaries,and exploration objects; 3) An analysis module, based on spatial interpretation, calculates spatial occurrence planes and makes inference about three-dimensional spatial intersection lines. In terms of specific functions, the mobile terminal focuses on field acquisition, while the desktop terminal on indoor analysis; both cooperate to achieve a seamless connection between office work and field work. The system has been applied to the geological survey of a hydraulic project in western China, and it shows a significant improvement on geological recording efficiency and quality in comparison with the traditional survey.

文章以青山冲水库右岸溢洪道边坡为项目背景,利用FLAC3D软件对边坡在开挖及采取锚索、抗滑桩等支护措施后的变化规律进行数值模拟分析,分析了自然边坡、开挖边坡、支护边坡在正常工况与非正常工况下的应力应变及稳定情况,研究了次生软弱夹层R1对边坡的影响,提高了对边坡处理效果的把握和认识,为设计提供理论依据和支撑.

近年,信息技术在水利水电工程地质勘察中得以应用,在地质数据采集、存储、管理、分析、三维可视化展示等方面均有所发展.对国内外水利水电工程地质勘察应用新技术的情况进行了研究,分析了GIM、工程数据库、无人机倾斜摄影、GIS、GPS等技术与地质勘察的融合,简单介绍了中水北方勘测设计研究有限责任公司自主开发的"水利水电工程三维地质勘察系统",展望了勘察信息化未来的发展趋势.

近年来,随着智慧水利的发展,带动了相关行业信息化的不断提升.地质与测绘是传统专业,也是工程建设的基础学科,其信息化水平的提升,对水利工程的顺利展开至关重要.以智慧水利为切入点,针对地质与测绘技术现状问题,分析其信息化发展趋势,探讨在智慧水利中,勘测行业如何提升信息化能力,并在信息化时代发挥更大的作用.

During the fieldwork of hydraulic engineering, practical engineers normally document geological information manually. Although there are some GIS-based digital tools for geology, they are not perfectly applicable to hydraulic engineering. As a result, the current work mode is ineffective, unmanageable, error-prone, and not conducive to subsequent analysis. To address this problem, we developed a digital tool which enables geological recording and quick modeling based on 3D real scenes in the field of hydropower projects. There are three modules in the surface tool: object recording, image interpretation, and field analysis. The object recording module is to mark geological points (e.g., drills and shafts), lines (e.g., faults, stratigraphic boundaries), and surfaces (e.g., slope and stocking yard) on a 3D scene and then store them in the database. The image interpretation is to interpret the 2D information in images to 3D models loaded in 3D software for further studies, such as GOCAD. The field analysis includes surface fitting, stability analysis of blocks, occurrences calculating, rock recognition, and 69/sketching. The tool is helpful for recording data, drawing geological boundaries, and building a preliminary model in the geological survey.

结构面的随机特征是水工边坡岩体稳定性研究的关键,目前一般采用单变量或双变量的概率模型研究,难以实现多参数的联合分析.基于生成对抗神经网络(GAN)和集成学习策略,提出了一种可用于结构面多维参数联合模拟的E-WGAN算法.相比于传统方法,该算法能够准确建立参数间的相关关系,实现岩体结构面形态的精确描述和模拟.实验对一组包含三个参数(迹长、倾向、开度)的结构面进行分析,证实了传统方法的局限性;而E-WGAN算法可同时对多维参数联合建模,从而还原结构面的真实分布特征.通过建立离散裂隙网络表明,利用该算法模拟的数据生成的迹线图对出露面的还原度更高.此外,该算法可被推广至更高维度地质参数分析,应用前景广阔.

大型复杂三维地质BIM模型具有地形地层复杂、构造多、地质界面构网数据量大等特点,严重制约向CAE数值计算模型的转化.利用MicroStation的MDL、ANSYS的APDL、FLAC3D的Fish语言进行数据转化接口的二次开发,并结合GeoStation强大的三维地质构建功能,提出一种结合ANSYS优化GeoStation中地质界面网格质量并在ANSYS实现地质实体的建立及剖分,最终导入FLAC3D中进行数值计算的解决方案.将此方法应用于某一工程实际,大大提高了建立三维地质计算模型的效率.

水利水电工程地质勘察涉及内容多、专业配合紧密、作业流程繁杂,传统的生产技术与手段已严重落后,难以满足日益紧张的生产需求.文章立足于地质生产的全过程及全生产要素的应用需求,基于数字孪生技术提出了水利水电工程地质的数字化应用方案,包括研究背景、方案架构、与大数据等新一代信息技术的融合与应用、研究内容等,力图实现地质勘察的内外业一体化、云端一体化、天地空协同、智能决策分析、三维地质模型快速建立等功能,助力水利水电及其他行业勘察业务的转型升级.

某水电站地下厂房所处地质条件复杂,为全面展示地质对象的空间位置及相对关系,利用GeoStation建立了地下厂房区的三维地质模型,工作内容包括完善工程地质数据库、建立地质界面及地质异形体等.三维地质模型形象直观地展示了地下厂房所处的地质条件,为地下厂房设计起到了辅助支撑作用.

文章利用GOCAD的DSI插值技术,以工程区地形等高线为例,在GOCAD中插值重构了地形界面,分析了待建模区域等高线完整和部分缺失情况下的地形建模方法;通过MDL自编程序,将所得网格节点数据导入Mi-croStation快速形成Mesh面.Mesh面在保证精度的同时,节点数得以大大优化,较好的弥补了MicroStation“地形模型”模块在构建地质界面的不足,具有一定的工程参考价值.

大型水利水电项目大都处于高山峡谷,其地质构造复杂、地质信息众多,给工程勘测、设计与施工带来了极大的困难.传统的二维地质图表达方式已经很难满足工程空间分析的需求.拟用AglosGeo三维地质软件,结合新疆某大型水电站工程,针对坝址区复杂地质构造的三维地质建模技术和工程应用进行简要阐述.

With the rapid development of water conservancy industries ,in order to satisfy the increasing requires for qual-ity and efficiency, the importance of 3D geological modeling has become more and more prominent .The authors use GeoStation software based on MicroStation 3D platform,set a 3D geological model of Gelantan Hydropower Station ,Lix-ianjiang River,Yunnan,to give a brief introduction on modeling process and utilization .

随着科技进步,水电站设计领域的设计手段发生着日新月异的变化,由传统的二维平面设计向立体、直观、准确的三维设计转变是水电站设计发展的必然趋势.结合三维可视化协同设计工作,就水电站建筑相关专业三维模型的建立与布置、材料统计,以及从三维模型中剖切二维施工图纸等技术进行研究探讨,结合水电站厂房设计实践,对基于Bentley软件平台的水电站建筑、结构和设备等多专业三维协同设计的实现方法、流程及存在的主要问题进行了探索和总结.

This subject,based on water diversion project in central southern Ningxia,studies engineering properties of soft rock in soft rock tunnel and gives the scientific basis for engineering design,construction and safety operation. In the subject,some properties of weak surrounding rocks,including the hydro-physical properties,swell-shrink property,rheological properties,mechanical properties and deformation mechanism,have been studied by several methods of exploration,the laboratory tests,the in-situ tests,the genetic mechanism method. From composition and existing environmen of the soft rock,it reveals the difference of the soft rock property between areas and the deformation law of the weak surrounding rock with the change of environment. It also puts forward supporting time and supporting measures of the tunnel in the layered soft surrounding rock with gentle occurrence. Due to its its strong hydrophilicity,a series of argillaceous soft rocks of Tertiary and Cretaceous,which are widely distributed in central and southern Ningxia area,have the characteristics of swell-shrink,slaking,rheology,time effect and sensitivity to the environment and are liable to cause the deformation failure of the tunnel. In consideration of the principle of engineering safety and investment control,this research result puts forward effective measurements that can be taken to inhibit or reduce the adverse impact on the mudstone engineering from excavation method,protection time and supporting technology.

Though the foundation rocks of Gomal Zam Dam Project are mainly composed of hard limestone ,they cannot meet the design requirements for gravity dam due to their low deformation modulus and high fracture index as a results of jointing and faulting .The stability of right abutment against sliding is also cann 't meet the code requirement due to the ex-istence of fault F2,F3 and F13.Because the river valley is rather narrow and steep ,and the rocks at depth are still frac-tured,it is unsuitable for deep excavation .Thus,the dam has to be based on moderately stress-relieved rocks,which would be treated by complete consolidation grouting to improve the rock mass integrity and deformation modulus .Meanwhile the concrete replacement method is applied for treatment of Fault F 2 to enhance the overall stability of the right bank abut-ment.