基于BP神经网络麻阳地区滑坡稳定性分析与预测

发布时间：2018-03-18 17:21

本文选题：滑坡　切入点：BP神经网络　出处：《湖南科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：麻阳地区位于沅麻盆地中南部,处于雪峰山与武陵山脉之间,是湖南省滑坡地质灾害易发区之一。为了查明麻阳地区潜在滑坡地质灾害,科学、合理地采取防灾、减灾措施,保护人们生命财产安全,依托《湖南省麻阳苗族自治县1:5万地质灾害详细调查》项目,开展基于BP神经网络麻阳地区滑坡稳定性分析与预测的一系列研究工作,取得了如下研究成果:(1)针对麻阳地区滑坡地质灾害的发育特点,以“地质过程机制分析量化评价”的学术思想为核心,结合野外地质灾害详细调查,建立了一套滑坡稳定性分析与预测的方法,形成了较为完整的研究滑坡稳定性分析的技术路线。研究成果直接指导区域性地质灾害防治,取得了较好的应用效果。这套技术路线和方法的建立,丰富了滑坡稳定性分析与预测的研究内容,对类似地质条件下滑坡的稳定性分析具有一定的指导意义。(2)在系统分析研究麻阳地区滑坡资料的基础上,通过运用数理统计与数值分析方法,详细地研究了该地区滑坡灾害类型与分布特征。研究结果表明,区内滑坡地质灾害主要发生海拔高度为175m-275m的丘陵区,红层中滑坡占总数的88.37%,顺向滑坡占总数的44.96%。滑坡发育概率与坡度间关系近似服从参数为μ=33.1,σ2=10.22的正态分布,且当坡度为23°~43°时,发生滑坡的概率最高。(3)在对区内滑坡灾害类型及分布特征研究的基础上,筛选出平均坡度、降雨强度等八大滑坡致灾因子,建立了适用于天然和暴雨两种不同工况下的麻阳地区BP神经网络滑坡稳定性分析与预测模型,并验证了七处边坡稳定性,其准确率较高,证明所建立的预测模型具有参考意义。(4)通过运用Geo Studio软件,对麻阳地区典型土质边坡的稳定性进行了数值模拟研究,证明了应用Geo Studio软件中的Slope/W模块对麻阳地区土质边坡稳定性分析是可行。(5)分析探讨了赤平投影稳定性分析方法的优缺点。推导论证后,发现公式适用于仅受重力作用的岩质边坡平面与楔形两种不同变形破坏模式。通过系统的对麻阳地区滑坡地质灾害类型及分布特征研究,以及基于Matlab的BP神经网络在麻阳地区滑坡稳定性分析与预测中的研究,构建了适用于麻阳地区的滑坡稳定性分析与预测模型,其将对该地区滑坡地质灾害的防治具有指导作用。同时,也为其它从事地质灾害研究工作者提供新的研究基础和理论依据。
[Abstract]:The Mayang area is located in the central and southern part of the Yuanma Basin, between Xuefeng Mountain and Wuling Mountains, and is one of the prone areas for landslide geological disasters in Hunan Province. In order to find out the potential landslide geological hazards in Mayang area, scientific and reasonable measures for disaster prevention and mitigation should be taken. To protect the safety of people's lives and property, and to carry out a series of research work on the analysis and prediction of landslide stability in Mayang area based on BP neural network, relying on the project of "1: 50,000 geological hazard detailed investigation" in Mayang Miao Autonomous County, Hunan Province. The following research results have been obtained: (1) according to the developmental characteristics of landslide geological hazards in Mayang area, the academic thought of "quantitative evaluation of geological process mechanism" is taken as the core, and combined with the detailed investigation of field geological hazards, A set of methods for landslide stability analysis and prediction are established, and a relatively complete technical route for landslide stability analysis is formed. The research results directly guide the prevention and treatment of regional geological hazards. The establishment of this set of technical routes and methods enriches the research contents of landslide stability analysis and prediction. On the basis of systematic analysis and study of landslide data in Mayang area, the method of mathematical statistics and numerical analysis is used to analyze the stability of landslide under similar geological conditions. The types and distribution characteristics of landslide disasters in this area are studied in detail. The results show that the landslide geological hazards mainly occur in hilly areas with an altitude of 175m-275m. In the red bed, the landslide accounts for 88.37% of the total, and the direct landslide accounts for 44.96% of the total. The relationship between the development probability of the landslide and the slope is approximately applied to the normal distribution with parameters 渭 33.1, 蟽 2 and 10.22, and when the slope is 23 掳or 43 掳, The probability of landslide occurrence is the highest. (3) based on the study of the types and distribution characteristics of landslide disasters in the region, eight major landslide disaster factors, such as average slope, rainfall intensity, etc., are screened out. The stability analysis and prediction model of BP neural network landslide in Mayang area is established, which is suitable for natural and rainstorm conditions. The stability of seven slopes is verified, and the accuracy of the model is high. It is proved that the prediction model has reference significance. (4) by using Geo Studio software, the stability of typical soil slope in Mayang area is numerically simulated. It is proved that it is feasible to use the Slope/W module of Geo Studio software to analyze the stability of soil slope in Mayang area. It is found that the formula is suitable for two different deformation and failure modes of rock slope which are only affected by gravity. The types and distribution characteristics of landslide geological hazards in Mayang area are systematically studied. And the research of BP neural network based on Matlab in the stability analysis and prediction of landslide in Mayang area is carried out, and a landslide stability analysis and prediction model suitable for Mayang area is constructed. It will play a guiding role in the prevention and treatment of landslide geological hazards in this area. At the same time, it will also provide new research basis and theoretical basis for other researchers engaged in geological hazard research.
【学位授予单位】：湖南科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P642.22

【二级参考文献】