用概率神经网络对水电边坡稳定性预测

概率神经网络(PNN)是一种训练速度快、结构简洁明了、应用广泛的人工神经网络,该方法采用贝叶斯分类决策理论建立系统的数学模型,以高斯函数作为激励函数,具有非线性处理和抗干扰能力强等特点。阐述了PNN的基本结构及其训练算法,提出了基于PNN的水电边坡稳定性预测方法,并采用免疫进化算法(IEA)对高斯型函数的标准偏差σ进行参数优化。理论分析和实例结果验证了基于IEA-PNN的水电边坡稳定性预测方法的正确性和有效性。

水电岩质边坡的稳定性概率分级方法研究

引入荷兰学者R.Hack等[1-3]提出的边坡稳定性概率分级(SSPC)方法,针对该方法在水电边坡稳定性评价中存在的两个局限性:一是仅适合于45 m以下边坡的稳定性评价和对完整岩石抗压强度的估计主观性和随意性较强,提出采用Hoek-Brown经验强度准则和边坡最大坡高经验公式,对SSPC方法中边坡岩体的抗剪强度和最大坡高进行修正。典型水电边坡实例的研究结果表明,SSPC修正方法对34个水电边坡稳定性评价的准确率为61.8%,对10个非构造性控制破坏的水电边坡稳定性评价的准确率达80%。SSPC修正方法对于水电岩质边坡的稳定性概率分级具有较好的适用性,可以为水电岩质边坡的稳定性快速评价提供一条新的途径。

锅浪跷水电站大型高边坡工程设计研究

锅浪跷水电站边坡工程地质条件复杂,边坡工程加固难度大、等级高,具有一定的特殊性。文章根据各建筑物的布置和地质地形条件,并结合岩体质量分类,进行边坡开挖和支护处理设计研究。对大坝坝肩及趾板、进水口后坡和厂房后边坡等重要部位的工程边坡,采用岩质边坡稳定性分析系统,详细地进行了不同段边坡的稳定分析计算和支护设计,为工程的顺利实施提供了技术支持。

工程边坡应用Hoek-Brown强度准则时的扰动因子取值探讨

Hoek-Brown强度准则在过去的20 a里得到不断的更新和完善。最近Hoek和Brown对以往的版本进行了归纳和总结,提出了2018版。这一版本纳入了定量的扰动因子D来确定强度指标,但是扰动因子D取值范围为0~1,涵盖了地下工程和边坡工程,又包括了从控制良好到控制不良的各种开挖工况,取值范围较大。对开挖爆破通常进行严格控制的水利水电工程边坡,需要研究其相应的扰动因子取值范围。对一个典型算例在大范围变动几何和物理力学参数情况下,比较了Hoek-Brown强度准则GSI—1995和GSI—2018两个版本的稳定安全系数。结果表明:当扰动因子D在0.78~0.88范围内取值时,两个版本求得的安全系数相当。本研究进一步考察了昌马、紫坪铺和柴石滩水库这3个曾按“1995版”开展过边坡稳定分析的工程实例,结果表明2个版本等效的扰动因子D取值范围为0.72~0.85。因此,在水电边坡工程分析中使用GSI—2018,建议其扰动因子取值范围可考虑在0.7~0.9之间。

三维激光扫描技术在边坡地质编录中的应用

西南地区地形条件复杂,高陡边坡发育普遍,其显著特点是地形起伏大,山体险峻。传统的地质编录方法是通过罗盘+皮尺进行现场地质测绘,不仅工作量大、成果精度低,而且存在人身安全隐患。以某水电站边坡为例,阐述了三维激光扫描技术在地质编录中的应用方法。其借助非接触式测量,能够快速高精度获取边坡实体密集点云数据,通过点云数据就能够解译边坡长大结构面和节理裂隙发育规律及空间展布特征,包括各级结构面分布、优势产状、延伸长度、间距等地质要素信息,从而研究结构面组合关系及其破坏模式,克服了传统地质编录的局限性。其地质要素的解译成果具有信息量丰富、准确度高的特点,将复杂的野外工作转化为内业数据进行处理,大大提高了工作效率,颠覆了传统的调查方法,具有一定创新性和先进性。

岩土工程监测信息管理与数据分析网络系统开发及应用

针对目前水电边坡工程海量的监测信息和监测数据,为减轻监测信息管理与数据分析的劳动强度,提高劳动生产率,采用Microsoft Visual C++网络编程技术和SQL Server网络数据库开发出具有先进性、可靠性、通用性和可扩充性的岩土工程安全监测信息管理与数据分析远程可视化网络系统。将该系统应用于龙滩水电站大坝左、右岸边坡及导流洞监测资料的分析处理,实现了监测信息和数据的远程实时共享及网络化的管理和分析,使监测分析成果能够及时、准确地反馈给设计人员,对规避设计和施工风险、保证边坡施工和运行安全起到了重要的作用。

工程地质科技发展与进步

进入21世纪,工程地质勘察新技术、新方法发展迅猛,西北院为适应日益竞争激烈的市场,以科技为引领,提高勘察技术进步。近年勘察技术进步主要在3D激光扫描技术、工程勘察技术数字化、定向钻孔测试设备研发及综合勘察技术、地质遥感解译、岩土数值模拟等方面。科技研发主要集中在复杂边坡、深厚覆盖层、岩体结构、特殊性岩土等方面,在以下方面取得研究进展:层状岩质边坡复杂变形破坏机理、失稳模式及治理措施;水库滑坡及塌岸稳定分析、安全防控措施研究及其工程应用;倾倒岩体变形破坏与分类;工程岩体缓倾角结构面;基于量化指标的高拱坝建基岩体可利用标准和建基面选择;超高面板坝巨厚砂砾石料层“深井法”勘察与可利用性评价技术;深厚覆盖层坝基工程特性;城市地下铁道工程富水砂卵石地层勘察与评价;西北地区大型风力发电场盐渍土工程地质特性等。

Interaction between anti-shear galleries and surrounding rock in the right-bank slope of Dagangshan hydropower station

The fight-bank slope of the Dagangshan hydropower station located in Southwest China is a highly unloaded rock slope. Moreover, large-scale natural faults were detected in the slope body; some excavation-induced unloading fractures were discovered at elevations between lo75m and 1146m. Because of poor tectonic stability, the excavation work was suspended in September 2009, and six large- scale anti-shear galleries were employed to replace the weak zone in the slope body to reinforce the fight- bank slope. In this study, based on microseismic- monitoring technology and a numerical-simulation method, the stabilities of the slope with and without the reinforcement are analysed. An in-situ microseismic-monitofing system is used to obtain quantitative information about the damage location, extent, energy, and magnitude of the rocks. Thus, any potential sliding block in the fight-bank slope can be identified. By incorporating the numerical results along with the microseismic-monitoring data, the stress concentration is found to largely occur aroundthe anti-shear galleries, and the seismic deformation near the anti-shear galleries is apparent, particularly at elevations of 121o, 118o, 115o, and 112om. To understand the interaction mechanism between the anti-shear gallery and the surrounding rock, a 2D simulation of the potential damage process occurring in an anti-shear gallery is performed. The numerical simulation helps in obtaining additional information about the stress distribution and failure-induced stress re-distribution in the vicinity of the anti-shear galleries that cannot be directly observed in the field. Finally, the potential sliding surface of the right-bank slope is numerically obtained, which generally agrees with the spatial distribution of the in-situ monitored microseismic events. The safety factor of the slope reinforced with the anti-shear gallery increases by approximately 36.2%. Both the numerical results and microseismic data show that the anti-shear galleries have a good reinforcement effect.

New developments of technology in rock engineering in China

In terms of rock engineering and technology in hydropower construction,the slope stability and monitoring techniques for high slopes of Three Gorges Project,the stability and support technology for high slopes of hydropower projects in deep river valley,the stabilization techniques for underground cavern group with large span and high side walls are introduced in this paper.As for rock engineering and technology in highway and railway construction,the Qinghai-Tibet Railway — new construction techniques in permafrost,the support techniques for large squeezing deformation in Wuqiaoling Tunnel,the construction techniques for tunnels in alpine and high-altitude region,the geological prediction techniques for tunnels in karst region,the microseismic monitoring and early warning techniques for rockbursts in deep and long tunnels are presented.For rock engineering and technology inmining engineering,the innovative techniques for roadway support inmines,the simultaneous extraction technique of pillarless coal and gas in coal seams with low permeability,the safe and efficient deep openmining technology,advances in monitoring,early warning and treatment ofmine dynamic disasters are discussed.In addition,the new anchorage techniques and precision blasting technique in rock engineering are introduced.

High Slope Stability of Diversion Power System Intake of Jinchuan Hydropower Station

The excavated height of the left bank slope of the diversion power system intake in Jinchuan hydropower station is about 16o m. The stability and safety of the slope during construction and its operation/utilization become one of the most important geological engineering problems. At the same time, it is also crucial to select a safe and economic excavation gradient for the construction. We studied the problem of how to select a safe and economic slope ratio by analyzing the geological condition of the high slope, including the lithology, slope structure, structural surface and their combinations, rock weathering and unloading, hydrology, and the natural gradient. The study results showed that the use of an excavation gradient larger than the gradient observed during site investigation and the gradient recommended in standards and field practice manuals is feasible. Then, we used the finite element method and rigid limit equilibrium method to evaluate the stability of the excavation slope under natural, rainstorm and earthquake conditions. The calculated results showed that the excavated slope only has limited failure, but its stability is greatly satisfactory. The research findings can be useful in excavation and slope stabilization projects.

Distribution and quantitative zonation of unloading cracks at a proposed large hydropower station dam Site

Rock mass unloading is an important rock engineering problem because unloading may impact the stability of a rock mass slope. Based on hydroelectric engineering principles, this study focuses on the classification of unloading zones to reflect the rock mass structure characteristics. Geological background and slope structure of the study region were considered to investigate the distribution and deformation of the unloading process. Quantitative indices were classified according to the formation mechanisms and the geological exhibition of unloading zones. The P-wave velocity(V_P), the ratio of the wave velocity(V_p) the ratio of the test P-wave velocity along the adit depth to the P-wave velocity of intact rock, the sum of joint openings every 2 meters(S_t), and the density of open joints(D_t) were calculated as quantitative indices for the rock mass unloading zone. The characteristics of the unloading zone of rock mass slopes at the dam site were successfully determined. The method of combining qualitative data with quantitative indices was found to be effective for the classification of slope unloading zones.