陕北公路边坡地质灾害危险性定量评价

发布时间：2019-05-05 09:18

【摘要】：陕西省是西北地区的交通枢纽城市,省内的高速及干线公路连通广大西部地区。陕北地处特殊的黄土高原地区,气候季节性差异较大,地形复杂,常发生多种公路边坡地质灾害,且损失较严重。因此对陕北公路边坡地质灾害(崩塌、滑坡)的危险性,进行定量评价很有必要。在总结国内外学者和经实地调研的基础上,依托陕西省干线公路“灾害数据库”资料,分析灾害发生的特征和规律。针对地质条件、几何特征、环境因素、破坏特征和其他特征这五个影响因素,组建了由17个指标组成的评价体系。结合SPSS20.0软件技术,以陕北为例,先对各定性评价指标重新编码,再进行K-均值聚类分析,将隐患边坡分为6类。选择类型2和类型3的隐患边坡,进行因子分析,得到公因子的计算模型及其权重。从这两大类型中分别抽取90%的边坡数据,经多次试算;针对每一类别,由因子旋转矩阵,得到从指标体系中提取出较为独立的6个公因子;由因子得分系数矩阵,得出6个公因子计算模型;并将公因子对应的方差贡献率作为权重。建立类型2和类型3各自的评价模型并进行检验。由模型计算出参与建模的90%边坡的危险值,经归一化处理后,依据危险值累积频率柱状图的特征和经现场确认,将两个类别统一的分为五个危险等级,并确定分级范围,具体分为:极高(0.9,1.0)、高度(0.7,0.9)、中度(0.5,0.7)、低度(0.3,0.5)和极低危险(0.0,0.3)。最后对剩余未参与建模的10%边坡的危险性,代入模型评价,并经实例验证,对照实际的边坡危险状况,评价结果与实际相符。本研究为陕西省公路边坡的灾害预警预测提供指导,为进一步的风险评价奠定基础,是定量性评价方法的一次有益尝试。
[Abstract]:Shaanxi Province is a transportation hub city in Northwest China. The highway and trunk roads in the province connect the vast western region. North Shaanxi is located in the special loess plateau region, the climate is different, the terrain is complex, many kinds of highway slope geological disasters often occur, and the loss is serious. Therefore, it is necessary to quantitatively evaluate the risk of slope geological hazards (collapse, landslide) in northern Shaanxi Province. Based on the summary of domestic and foreign scholars and on-the-spot investigation, the characteristics and rules of disaster occurrence are analyzed on the basis of the "disaster database" data of Shaanxi Province trunk roads. According to the five influencing factors of geological conditions, geometric characteristics, environmental factors, destruction characteristics and other characteristics, an evaluation system consisting of 17 indexes was established. Combined with SPSS20.0 software technology, taking the north of Shaanxi as an example, the qualitative evaluation indexes are re-coded first, then K-means clustering analysis is carried out, and the hidden slope is divided into 6 categories. The hidden danger slope of type 2 and type 3 is selected and the calculation model of common factor and its weight are obtained by factor analysis. 90% of the slope data are extracted from these two types and calculated several times. According to each category, six independent common factors are extracted from the index system by the rotation matrix of factors. Based on the factor score coefficient matrix, six common factor calculation models are obtained, and the variance contribution rate corresponding to the common factor is taken as the weight. The evaluation models of type 2 and type 3 were established and tested. The hazard values of 90% of the slopes involved in modeling are calculated by the model. After normalization, according to the characteristics of the cumulative frequency histogram of the hazard values and the field confirmation, the two categories are uniformly divided into five hazard grades and the classification range is determined. The results were as follows: extremely high (0.9, 1.0), height (0.7,0.9), moderate (0.5,0.7), low degree (0.3,0.5) and extremely low risk (0.0,0.3). Finally, the risk of the remaining 10% slope which is not involved in the modeling is evaluated by the model, and verified by an example, the evaluation result is consistent with the actual situation of the slope hazard. This study provides guidance for early warning and prediction of highway slope disaster in Shaanxi Province and lays a foundation for further risk assessment. It is a useful attempt of quantitative evaluation method.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U416.14

【参考文献】