基于人工神经网络的地裂缝危险性评价系统

Fatalness assessment system of earth fissure based on artificial neural network

现代地裂缝在世界许多国家普遍存在 ,已成为当今世界范围的主要地质灾害之一。利用地理信息系统 (GIS)与人工神经网络 (ANN)耦合技术建立了地裂缝灾情非线性模拟评价系统。作者在分析地裂缝灾害成因的基础上 ,利用地理信息系统 (GIS)的空间分析功能 ,建立了构造、地下水开采、地层和地貌 4个地学信息专题层图 ;采用人工神经网络 (ANN)这一以工程技术手段模拟人脑神经网络的结构和功能特征的技术系统 ,建立了地裂缝灾害危险性非线性模拟评价模型 ,开发研制了危险性评价系统 ,进而对榆次地裂缝灾害危险性进行了非线性模拟评价 ,将研究区按危险性系数进行了分区 ,为榆次城建。

Modern earth fissure appears in almost every country recently and it has been a main kind of geo hazard up to the present day.Based on the coupling technolgy of GIS & ANN,this article has developed the nonlinear simulation and assessment system of earth fissure.On the basis of analyzing the causes of earth fissures,this article setups four geo information layers (such as structure,groundwater mining,stratum and physiognomy) with the spatial analyzing function of GIS.And this article setups the nonlinear simulation and assessment model of earth fissure with ANN,the engineering technology of simulating the neural net of mankind.And with this assessment system,this article evaluates the fatalness of earth fissure in Yuci,Shanxi,plots the fatal grade according to fatalness coefficient,and provides the scientific basis for the decision making of city construction,environmental protection and land programming in Yuci City.