论文部分内容阅读
滑坡的易发性是对区域滑坡发生的一种空间预测,是滑坡灾害自然属性特征的体现,它与孕灾环境的各项因子密切相关。在Elman神经网络模型的基础上,尝试将其与灰色系统关联分析方法结合,利用灰色关联分析方法来客观地选出滑坡的主要影响因子,并将这些因子作为Elman神经网络模型的输入变量,也解决了Elman神经网络输入变量的优化选择问题。本文通过两个不同的研究区(福建闽清闽侯两县和建宁县)发生的滑坡点对Elman神经网络进行训练拟合,最后利用拟合好的模型对整个研究区的滑坡进行易发性空间预测,并获得理想的预测结果。
The vulnerability of landslides is a kind of spatial prediction of landslides. It is a manifestation of the natural attributes of landslide hazards. It is closely related to various factors of the disaster environment. Based on the Elman neural network model, this paper tries to combine it with the gray system correlation analysis method, and uses the gray relational analysis method to objectively select the main influence factors of the landslide, and takes these factors as the input variables of the Elman neural network model. Solve the Elman neural network optimization of input variables selection problem. In this paper, Elman neural network is trained and fitted by landslide points in two different study areas (Fujian Minhou County and Jianning County). Finally, the fitted models are used to carry out the landslide susceptibility Spatial prediction, and obtain the ideal prediction result.