摘要
本文以广西贺州市平桂区为例,以遥感解译资料为基础,利用GIS空间分析功能对研究区数据进行提取分级、赋值统计及归一化等处理,构建了包括坡度、坡向、地层岩性、植被指数、断裂构造、路网密度、土壤含水率、地形曲率、居民地密度的地质灾害易发性评价指标数据集;通过粗糙集理论及遗传约简算法对地质灾害易发性评价因子进行筛选,最后建立基于BP神经网络的崩塌滑坡易发性预测模型。BP神经网络模型与GIS技术相结合预测崩塌滑坡的易发性具有可行性。
Taking Pinggui district,Hezhou,Guangxi Province as an example,based on remote sensing interpretation,the GIS spatial analysis function was used for data extraction and classification assigned statistics,and normalized processing in the study area,while the geological hazards evaluation index data set were constructed including slope,slope direction,formation lithology,vegetation index,fault structure,road network density,soil moisture content,terrain curvature,and resident density.Geological hazard vulnerability evaluation factor were selected by rough set theory and genetic reduction algorithm.Finally,this paper built a predictive model of landslide vulnerability based on BP neural network.It is feasible to predict the vulnerability of geological disasters by the combination of BP neural network model and GIS.
作者
吴晶晶
江思义
吴秋菊
李海良
邱恩露
WU Jingjing;JIANG Siyi;WU Qiuju;LI Hailiang;QIU Enlu(Nanning Natural Resources Information Group Co.,Ltd.,Nanning 530021,Guangxi,China;The Guangxi Zhuang Autonomous Region Geological Environment Monitoring Station,Nanning 530029,Guangxi,China;Remote Sensing Center of Guangxi,Nanning 530023,Guangxi,China)
出处
《资源信息与工程》
2021年第4期100-104,107,共6页
Resource Information and Engineering
基金
广西壮族自治区国土资源厅地质调查项目(桂国土资办[2018]317号)。
关键词
GIS
BP神经网络
地质灾害
易发性预测
GIS
BP neural network
geological hazards
vulnerability prediction
作者简介
吴晶晶(1994-),女,广西北海人,本科,助理工程师,主要从事水文与工程地质相关工作。
