【作者】 朱济祥；

【导师】 练继建；

【作者基本信息】 天津大学 ， 水利水电工程， 2007， 博士

【摘要】 李家峡坝址区工程地质条件十分复杂,其中左岸f₂₀和F₂₆,右岸F₂₇等断层的稳定状况是影响李家峡水电站安全运行的重要因素。本课题以坝肩附近存在地质缺陷的黄河李家峡大坝为背景,根据李家峡大坝监测系统采集的数据,建立大坝安全监控模型和指标,对大坝和基础的工作性态和安全状况进行实时、综合的评价,实施分级报警,以确保电站枢纽的安全运行。（1）利用大型有限元软件ANSYS建立了规模较大、模拟介质类型较全面的包括李家峡拱坝及周围一定范围地基的三维有限元模型,对李家峡大坝开挖、坝基（肩）表层和深层基础处理后以及大吨位预应力岩锚群实施后等阶段不同荷载工况下的坝肩稳定性进行了全过程仿真分析。对大吨位预应力锚索群加固的力学机理及在提高坝肩安全储备方面的作用进行了较深入的分析;（2）收集、整理、分析拱冠、左、右1/4拱的垂线、库水位、气温及拱冠断面不同高程处的坝体温度等监测资料;（3）利用确定性模型计算大坝在不同参数、工况下各高程部位的变位值,验证了大坝的安全运行是有保障的;（4）利用人工神经网络对大坝及基岩的力学参数（Ec、Er）进行反演计算,求出实际参数值;（5）根据现有的原型观测资料建立坝体变位的分别基于大气温度和坝体实测温度的统计模型及混合模型,各模型的精度指标满足要求,且利用反演参数建立的混合模型用作监控模型更可行;（6）利用BP神经网络建立了大坝变形多测点监控模型,拟合程度较高;（7）利用改进的遗传算法建立了大坝变形的遗传回归监控模型,以进一步改进预测效果;（8）尝试建立基于小波分析的监控模型;（9）据结构计算分析法,拟定了李家峡大坝一、二、三级监控指标,为大坝的安全运行决策提供了重要的参考价值。更多还原

【Abstract】 The engineering geologic conditions within Lijiaxia Arch Dam area of the Yellow River are quite complicated , of which the stability of faults such as f20 and F26 in the left bank ,as well as F27 in the right bank is one of the important factors that affects the safe operation of the hydro-electric station. So the Lijiaxia Arch Dam on the Yellow River is used as the prototype in this dissertation. Based on the data collected from the dam’s monitoring system, the aim of the research is to establish the project’s monitoring models and monitoring indexes, and to make a real time and comprehensive evaluation on the working and safety condition of the dam and its foundation, as well as make stage alarm, so as to ensure the hydro complex’s safe operation.By use of the finite element software of ANSYS, a 3-D finite element model has been established, which encompasses the arch dam and its surrounding region. And many of the materials have been simulated. The analyses of whole process simulation of the abutment stability under different load levels have been made, which include the procedures of excavation, foundation treatment in shallow and deep layers, as well as application of high-tonnage prestressed anchorage cables. The comprehensive and detail analyses of the mechanical mechanism of application of the high-tonnage prestressed anchorage cables and their effects on enhancement of abutment’s safety have been made.The monitoring data ,such as plumb-line displacements of the dam’s arch crown, the left and right one fourth arch of the dam, water levels of the reservoir, air temperature, temperatures at different levels of the dam’s arch crown, etc. have been collected and analyzed.By use of the deterministic model, displacements at different levels under different parameters and loading conditions have been calculated, which show us that the dam’s safe operation can be ensured.By use of the artificial neural network, the back calculation of the parameters of the dam and its foundation has been finished. And the real parameters have been obtained.In accordance with the known prototype monitoring data, two kinds of statistical models, which are separately based on the air temperatures and the dam’s monitored temperatures, and the hybrid model have been set up. And it is more practical to use the hybrid models which have been established by use of the back calculated parameters as monitoring models.The multi-points monitoring models of the dam’s deformation have been set up by use of BP neural network, which are of high fit goodness.The genetic regressive monitoring models of the dam’s deformation have been set up by use of the improved Genetic Algorithms (GA), so as to improve the precision of prediction.The monitoring models which are based on Wavelet analysis have been set up.By use of the analytical method of structural calculation, monitoring indexes at three different levels have been presented, which can give great help to the safety decision of the dam.更多还原