【作者】 孟凡奇；

【导师】 李广杰；

【作者基本信息】 吉林大学 ， 地质资源与地质工程， 2011， 博士

【摘要】 本文以“辽宁省鞍山市重点地区地质灾害隐患查证”项目为依托,在野外调查和室内资料整理工作的基础上,构建了泥石流三维地理信息数据库。基于COM的GIS组件ArcObjects9.2进行二次开发,以Visual Basic作为开发工具,开发了泥石流预测预报系统。在开发平台上,对泥石流流域的钻孔数据、地层、水系、泥石流堆积扇进行三维实体建模,实现了三维可视化模拟。选取了山坡坡度、相对高差、植被覆盖率、沿沟松散物储量、5天累计降雨量、最大小时雨强和当日雨量为泥石流预警的影响因子,构建了地质环境条件和降雨条件相结合的岫岩县泥石流预警指标体系,并运用改进的AHP法对影响因子权重赋值。利用功效系数法和可拓学理论对岫岩县泥石流沟进行了预警研究,开拓了基于泥石流沟的地质环境条件和降雨诱发条件为影响因子的非线性预警方法。在泥石流强度预测方面,运用逐步判别分析法筛选泥石流最大危险范围预测指标,利用非线性多元回归最小二乘法拟合来确定最大危险范围预测模型,对危险范围的最大长度、最大宽度、最大厚度、最大面积进行了预测。最后,在GIS平台上将泥石流沟预警和最大危险范围预测结果三维可视化输出,便于工作人员查询浏览。更多还原

【Abstract】 There are so many mountains in China. Influenced by geological structure, such as fracturation, numerous mountains are very steep, which possess rich unconsolidated accretion and low forest coverage. When rainy weather or even downpour appears, serious debris flow disaster will occur. The debris flow, occurring suddenly and overwhelmingly, has double effect on eboulement, landslide and floods, and also has huge destructive power which seriously imperils engineering facilities such as mountainous transportation, mine, water conservancy and hydropower, and safeties of lives and property of the local people in the mountains. Therefore, the reliability and validity of prediction of debris flow plays an important role in ensuring safeties of lives and property of the masses.At present, prediction of debris flow has been paid great attention by the domestic and overseas scholars and the local government. However, the overall level of prediction is not high and the accuracy is very low. The main reason for this condition is that most of existing methods, which neglect the influence of different geological environment condition on the occurrence of debris flow, early warn only according to the causative factor of rainfall or forecasting dangerous range of debris flow by perceptual knowledge on the basis of experience. How to alter debris flow with single gully, combining the geological environment information and the rainfall information together? And how to forecast debris flow dangerous range by adopting influencing factors of the geological environment? It is the key topic of debris flow prediction research.The research is conducted in this thesis on the basis of "Geological disaster verification of main areas in Anshan city of Liaoning province". As the key area of Anshan, Xiuyan county needs geological disaster verification, where the debris flow is the major geological disaster. After long-term crustal movement and exogenetic processes, the physiognomy of Xiuyan is low-middle mountains and hills zone. This region is the place where the faults develop well, the unconsolidated solid paricles are abundant, the effect of human activity is intense, the vegetation coverage is very low and the rainfall always appears in July or August. Because of the poor geological environmental quality, the debris flow occurs several times during the season of the rainstorm.The thesis introduces the divided graphic layer method of debris flow 3D GIS and the attributive structure of each layer based on GeoDataBase. The 3D graphics library consisting of contour line, stratum information, river system and debris flow basin, as well as debris flow GIS including debris flow feature database, disaster history database, rainfall database are built, which accomplishes spatial digitization of geological environmental information of debris flow to duly browse and inquire geological information data and provide database support for prediction task of debris flow.The method of building 3D model which is suitable for prediction of debris flow is raised on the basis of the knowledge of 3D data model. This thesis introduces the building method of debris flow 3D solid and successfully builds the 3D model of debris flow basin. The expression of drilling data as well as its building method is introduced and 3D layer model of debris flow is built in accordance with drilling data and IDW interpolation. Moreover, building method of 3D river system and debris flow deposit fan is introduced to accomplish 3D visualization and simulation analysis of debris flow.Based on GIS, the forecast model of debris flow is built by fully analyzing debris flow feature database, disaster history database, rainfall database, accomplishing the 3D prediction of debris flow early warning and maximal hazardous range and thickness of debris flow. Comprehensively considering the rainfall and geological conditions, hill slope, relative height, vegetation coverage, along the groove loose material reserves,5d cumulative rainfall, maximum hours of rainfall intensity and the date of rainfall are selected as the influencing factors. Using the optimal transitive matrix to improve the AHP to avoid the high-subjectivity in adjusting the judgment matrix in the later checkout process Using the improved analytic hierarchy process to calculate the weight coefficient of factors of debris flow warning. Using efficacy coefficient method and Extenics theory to establish early warning models of debris flow in Xiuyan county. In view of prediction of maximal hazardous range of debris flow, adopting stepwise discriminant analysis to select the influencing factors and using nonlinear multiple regression least square principle to predict hazardous range and accumulative thickness of debris flow. Delivering debris flow prediction results 3d visualization output on GIS platform, the powerful inquiry and the spatial analysis function will facilitate the comprehensive searching analysis of the results.The main algorithms adopted in this thesis are analytic hierarchy process, stepwise discriminant analysis, efficacy coefficient method, Extenics theory, least square method in data mining theory. On the basis of the Arc Objects 9.2—GIS component of COM and regarding Visual Basic as the development platform, the system of GIS debris flow prediction in advance is developed on the basis of GIS. This system consists of some modules, such as the management of spatial database, the construction of 3D model, spatial overlay analysis, spatial inquiry, geological forecast and prediction, etc. The functions of the system are the inputting, inquiring and modifying of debris flow characteristic data, rainfall data, disaster history data, drilling geophysical data, remote sensing data; 3D model building of drilling, strata, water system, and debris flow basin; spatial inquiring analysis characterized by clicking inquiry and location inquiry according to the specific attributes; forecasting module realized by the weight coefficient calculation to the influencing factors of debris flow early warning on the basis of debris flow prediction of improved analytic hierarchy process, efficacy coefficient method and Extenics theory; the debris flow prediction of nonlinear least squares multiple regression to the dangerous range and accumulative thickness and the 3D display of the dangerous range of debris flow.Adopting methods, theories and the developed system of debris flow prediction that have been described in this thesis, the debris flow in Xiuyan county is forecasted in two aspects of the debris flow early warning and prediction of maximal hazardous range and thickness of debris flow and several conclusions are obtained as follows:1. The poor quality of geological environmental in Xiuyan county is the fundamental factor of the occurrence of debris flow. The high-intensified rainfall is the only factor to cause debris flow.2. Based on field investigation and analysis results from predecessors, the factors of hill slope, relative height, vegetation coverage, along the groove loose material reserves,5d cumulative rainfall, maximum hours of rainfall intensity and the date of rainfall have a strong influence on the occurrence of debris flow in Xiuyan county. Therefore, the geological environmental condition and rainfall factors are combined as impact factors of early warning model.3. According to the contrastive analysis of forecast results between efficacy coefficient method and Extenics theory, the results of these two methods are basicly similar. It conforms to the reality and validates the forecasting ability of nonlinearity theory.4. By means of stepwise discriminant analysis, selecting the influencing factors of maximal hazardous range of debris flow, whose result is more objective. Debris flow dangerous range is divided into three groups by size. And adopting nonlinear multivariate regression least square method to forecast maximal hazardous range and thickness of three groups debris flow, the forecasting accuracy is very higher.5. The accuracy of forecasting results to debris flow early warning and maximal hazardous range and thickness, convincingly demonstrate that the debris flow prediction can be carried out by nonlinearity theory. The innovation of research thoughts and methods which has higher value of application and promotion, embodies that the selected subject and the research of this thesis has scientific significance and economic value.更多还原