人工神经网络在贵州地质灾害易发性分析中的应用探索

党杰; 陆安良; 李蕊; 王成龙

党杰,陆安良,李蕊,王成龙.人工神经网络在贵州地质灾害易发性分析中的应用探索[J].贵州地质,2024,(1):67-74

人工神经网络在贵州地质灾害易发性分析中的应用探索

Application Exploration of Artificial Neural Network in the Analysis of the Susceptibility of Geological Disasters in Guizhou

投稿时间：2023-08-16

DOI：

中文关键词: 人工神经网络地质灾害易发性分析贵州

英文关键词:Artificial neural network Geological hazard Susceptibility analysis Guizhou

基金项目:贵州省地质灾害综合遥感识别中心建设、探索开展地质灾害早期识别现场验证项目资助。

作者	单位
党杰	贵州省地质环境监测院，贵州贵阳550018
陆安良	贵州省地质环境监测院，贵州贵阳550018
李蕊	北京云阈科技有限公司，北京海淀100080
王成龙	北京数识科技有限公司，北京丰台100071

摘要点击次数: 154

全文下载次数: 429

中文摘要:

贵州省地质环境条件复杂、地质灾害发育严重，开展基于人工神经网络算法的地质灾害易发性分析，对提高全省易发性评价效率、实现智能化分析具有重要意义。本文通过对各种人工神经网络算法的比较，选取径向基神经网络(RBF)、概率神经网络(PNN)、模糊神经网络(FNN)作为模型算法，采用基于GIS平台的空间数据建模软件(SDM)构建基于人工智能的地质灾害风险分析模型系统，通过数据准备、数据处理、模型训练、模型调用、评估优化等步骤，开展基于三种人工神经网络的贵州山区地质灾害易发性分析应用探索。结果表明:(1)三种人工神经网络计算的易发性结果分区合理、精度检验合格，AUC检验显示具有良好预测价值；(2)通过与专家经验分析结果对比，RBF算法的评价结果与实际更为吻合，表明RBF算法可更好地应用于地质灾害易发性分析。

英文摘要:

The geological environment conditions of Guizhou Province are complex and geological hazards are seriously developedTherefore，it is of great significance to carry out geological hazard susceptibility analysis based on artificial neural network algorithm to improve the efficiency of the province’s susceptibility evaluation and realize intelligent analysisBy comparing various artificial neural network algorithms，radial basis function network(RBF)，probabilistic neural network(PNN)and fuzzy neural network(FNN)are selected as model algorithms，and spatial data modeling software(SDM)based on GIS platform is used to construct an artificial intelligence-based geological disaster risk analysis model systemThrough the steps of data preparation，data processing，model training，model invocation，evaluation and optimization，the paper carries out three kinds of geological disaster susceptibility analysis based on artificial neural networks in mountainous areas of GuizhouThe results of the three kinds of artificial neural network calculation are reasonably partitioned，the accuracy test is qualified，and the AUC test shows that it has good prediction value；Compared with the expert experience analysis results，the evaluation results of RBF algorithm are more consistent with the reality，which indicates that RBF algorithm can be better applied to the geological disaster susceptibility analysis

查看全文查看/发表评论下载PDF阅读器