斜坡单元支持下区域泥石流危险性AHP-RBF评价模型

通过分析目前区域泥石流危险性评价研究中评价单元、指标体系及其提取方法、评价模型3个关键内容对评价结果的影响，将斜坡单元作为区域泥石流灾害评价的基础评价单元并提出斜坡单元的自动定量划分方法；提出区域泥石流危险性评价的指标体系（即斜坡面积、主沟长度、斜坡相对高差、平均坡度、斜坡切割密度、平均主沟坡降、植被覆盖率、地层岩性共8个指标），并通过GIS技术提取指标体系；提出区域泥石流危险性评价的APH-RBF神经网络评价模型.以湖北省神农架林区灾害防治示范区木鱼镇为研究对象，采用已查明的泥石流灾害点的位置资料，验证本模型评价结果的合理性，为示范区的防灾减灾及村镇规划提供参考.
Abstract: The slope unit was presented taking as the basic assessment unit and automatic quantitative zoning method of slope unit was offered by analyzing the influence of the three important indexes:assessment unit, index system and its extraction methods, assessment models on the hazard assessment of regional debris flow. Index system for regional debris flow hazard assessment was put forward, including slope area, main gully length, slope surface relative relief, average slope, slope cutting density, main ditch slope ratio, vegetation coverage, formation lithology, which was extracted by GIS. APH-RBF neural network model was built to assess hazard of regional debris flow. The proposed method was applied in a slope analysis in Muyu, Shennongjia Forestry District, China. The results of the assessment model was verified by using the position information of debris flow points, which illustrates the reasonability of the proposed model that provides a reference for the demonstration area of hazard prevention and mitigation and town planning.