基于斜坡单元的地震滑坡敏感性分析

Susceptibility analysis of earthquake-induced landslides based on slope units

针对斜坡单元大小直接影响地震滑坡敏感性区划结果,论文利用河网密度优选出集水面积阈值,在此基础上生成最优斜坡单元。构建了基于遗传算法的支持向量机敏感性分区预测模型,并实现了宝盛乡地震滑坡敏感性分区。结果显示,在优选出的斜坡单元基础上完成的地震滑坡敏感性分析的精度达到了98.72%。利用优选斜坡单元结合基于遗传算法的支持向量机构建的地震滑坡预测模型是滑坡预测的有效工具,可为防灾减灾提供决策支持。

Slope Units size influences the result of earthquake-induced landslides susceptibility zoning directly. This article makes use of drainage density to optimize catchment area threshold, then generates the optimal slope unites in the basis of the best catchment area threshold, establishes support vector machine sensitivity prediction model based on genetic algorithm, implements seismic landslide susceptibility zoning of Baosheng Township. The results show that the accuracy of seismic landslide sensitivity analysis completed based on optimized slope element reaches to 98.72 per cent, and using optimized slope element combined with genetic algorithm-based support machine to es- tablish seismic landslide prediction model is an effective tool, and may provide a decision-making for disaster pre- vention and reduction.