山西省地质灾害气象预警模型探讨

Meteorological early warning model research geologic disaster in Shanxi province

山西省地质灾害气象预警模型的基本原理是采用系统论方法将地质灾害当作一个复杂的非线性动态系统进行研究,运用非线性动态组织系统模型、数据处理自组织模型、模糊信息优化模型、人工神经网络模型等数据预处理模型从孕灾环境、致灾因子和承灾体三个方面进行地质灾害风险性研究;选择地形、地貌、地层岩性、地质构造、水文流域、大气降水等六个单要素的致灾概率和承灾体的四个指标(人口、国内生产总值、工业总产值和农业总产值)进行专题研究,最终建立符合山西实际情况的基于信息扩散的模糊人工神经网络预警模型。该预警模型是动态的非线性自学模型,是一个"活"的模型,在近几年汛期地质灾害预警中起到了较好的预警效果。

The basic principle of meteorological forewarning model for geological hazards of Shanxi province is to take geologic hazard as a complex nonlinear dynamic system for research,by using nonlinear and dynamic system model,self-organized data processing model,fuzzy information optimization model,and artificial neural network model,etc. The geological disaster risk research is focus on three aspects,including hazard-formative environments,hazard-formative factors,and hazard-affected body. Six factors（ terrain,landform,formation lithology,geological structure,river basin,and meteoric waters） and four indicators of hazard-affected body（ population,GDP,total industrial output value,and total agricultural output value） are selected to research specifically. Eventually a fuzzy artificial neural network forewarning model,which is based on information diffusion and in line with practical situation of Shanxi Province,will be established. Such forewarning model is dynamic,nonlinear,and self-organized. It produces positive effect on geological hazard forewarning in recent flood seasons.