雅康高速—泸康段沿线泥石流危险性评价

发布时间：2018-01-24 16:30

本文关键词： 公路泥石流危险性评价遗传算法 BP神经网络　出处：《西南石油大学》2017年硕士论文　论文类型：学位论文

【摘要】：雅康高速是连接川西甘孜藏族自治州与内地的第一条高速公路,该高速公路将成为川西地区的交通枢纽。公路建设在四川盆地西部边缘与青藏高原东部边缘的过渡带,区内地形复杂、山高谷深、坡度较陡,对公路的建设和养护造成了极大的困难。公路穿越了龙门山断裂带、鲜水河断裂带、安宁河断裂带的交汇处,此区域内降水量大,地震活动频繁且地震烈度高,断裂活动极为强烈,人类经济活动增加,这些都为泥石流灾害的发育提供了极为有利的条件,对公路结构形成了巨大的威胁。深入研究雅康高速所在地区泥石流形成机理和发育特征,建立可靠的泥石流危险性评价体系,可以提供灾害评估和预警信息,对公路设施的保护和人民生命财产的安全的保障有着重要的科学意义和实际价值。本文通过对研究公路的实地考察和资料翻阅,探索研究区泥石流灾害的发生机理,确定对研究区泥石流灾害形成贡献较大的评价因子;将流域作为危险性评价单元,并以流域单元为单位提取各评价因子数据;通过遗传算法对BP神经网络进行优化,利用优化后的BP神经网络模拟研究区泥石流流域发生泥石流灾害的概率;在GA-BP神经网络的基础上建立针对研究区的泥石流危险性评价模型,最后实现对研究公路沿线的泥石流危险性评价。本文的主要研究工作如下:(1)以DEM为底图,结合研究区实际情况,利用ArcGIS10.2平台建对研究区进行流域划分。最终将研究区划分成集水面阈值大小分别为1km2、3km2、5km2和10km2的子流域。获取研究区子流域后再根据泥石流的运动特征,公路泥石流成灾形式以及Google Earth的三维真实地形叠加分析三个方面来筛选可能对公路造成影响评价单元。最后,确定的研究公路泥石流危险性评价单元有31个。(2)以研究区泥石流成灾条件为基础,构建研究公路泥石流评价因子体系,分别为:地形起伏度、平均坡度、断层密度、切割密度、地震烈度、年均降水量、地层岩性、年降水变差系数和归一化植被指数(NDVI)。本文将这9个因子以流域为单元进行了数据的提取和分析。(3)本文利用BP神经网络建立了泥石流危险性评价模型,并通过遗传算法对BP神经网络的初始权值和阈值进行优化,以此提高网络的性能。将经过训练的GA-BP神经网络的性能与未优化的BP神经网络进行对比,结果经过优化的BP神经网络的性能要高于未优化的BP神经网络,由此判定该网络模型适用于研究公路的泥石流危险性评价。(4)以划分好的流域单元对研究公路泥石流危险性进行评价,得出有1条流域为极度危险;高度危险的流域有12条;中度危险的流域有6条;轻度危险的流域有12条。
[Abstract]:Yakang high-speed is the first expressway connected Sichuan Tibetan Autonomous Prefecture of Ganzi and the mainland, the highway will become the Western Sichuan hub. Highway construction transition in the eastern edge of the Sichuan basin and the Tibetan Plateau on the western edge of the belt area with complex terrain, steep slope, deep Yamadaka Tani, caused great difficulties for the construction and maintenance of highway the highway through the Longmen Mountain Fault Zone, Xianshuihe fault zone, the intersection of the Anning River fault zone, the precipitation in this region is large, frequent seismic activity and earthquake intensity is high, fault activity is very strong, the human economic activity increases, these are for the debris flow development provided the most favorable conditions, has formed a huge threat to the highway structure. The further study of debris flow area high speed Yakang formation mechanism and development characteristics, to establish debris flow risk assessment system is reliable, can provide disaster Damage assessment and early warning information on the road infrastructure and protect people's lives and property safety protection has important scientific significance and practical value. Based on the research of highway fieldwork and data browsing, the mechanism of debris flow disasters in the study area, determine the formation of large contributions to the research on the evaluation factors of debris flow disaster area; the river as a risk assessment unit, and the basin unit to extract data of each evaluation factor; BP neural network optimized by genetic algorithm, probabilistic simulation study area of debris flow basin debris flow using the optimized BP neural network; establish the risk assessment of debris flow in the study area based on GA-BP neural model on the network, and finally realize the risk assessment of debris flow on the highway. The main research work of this paper is as follows: (1) using DEM as the base map, combining with the research The actual situation in the construction of watershed in the study area using the ArcGIS10.2 platform. In the end of the study area is divided into a set of water threshold size respectively is 1km2,3km2,5km2 and 10km2 sub basin. The study area after the sub basin according to the motion characteristics of debris flow, debris flow disaster type and Google Earth 3D terrain analysis in three aspects to screen the possible impact of the evaluation unit on the highway. Finally, the risk assessment of debris flow on the highway 31. The determination unit (2) in the study area of debris flow disaster conditions, construction of road debris flow evaluation factor system, respectively: relief, average slope, fault density, cutting density, seismic intensity. The average annual precipitation, lithology, the variation coefficient of annual rainfall and normalized difference vegetation index (NDVI). In this paper, the extraction of data and the 9 factors in Watershed Analysis. (3) this paper uses BP neural network to establish the debris flow risk assessment model, and the genetic algorithm of BP neural network to optimize the initial weights and thresholds, in order to improve the network performance. The performance of GA-BP neural network training method was compared with the BP neural network is not optimized, results after performance of BP nerve network optimization is higher than that of BP neural network is not optimized, thus determining the risk assessment of debris flow in the network model is applied to study the highway. (4) of the debris flow risk assessment of highway to unit watershed division well, obtains 1 rivers is extremely dangerous; in highly dangerous 12; there are 6 moderate risk River Basin; mild risk are 12.

【学位授予单位】：西南石油大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U418.56;P642.23

【相似文献】