基于空间信息格网与BP神经网络的灾损快速评估系统

A rapid evaluating system of flood disaster loss based on the spatial information grid and BP neural network

针对传统洪灾损失评估中存在的问题,建立了基于空间信息格网与BP神经网络的灾损快速评估系统.首先,利用Arc GIS空间分析技术提取洪灾区域,用信息格网技术将受灾区域格网化,分类提取不同受灾地类的受灾面积、淹没深度;其次,实地采集数据,构建受灾样本数据库;然后,建立基于空间信息格网与BP神经网络的灾损快速评估模型,并将评估模型使用程序编码的方式实现;最后,利用基于模型开发的评估系统针对某县受灾区域进行了洪灾经济损失的评估计算,结果表明,基于该模型方法计算结果与洪灾损失统计值相差不大,说明基于空间信息格网与BP神经网络的灾损快速评估系统具有一定的可靠性.

For traditional flood damage assessment problems, a rapid disaster damage assessment system based on spatial information grid and BP neural network is established in the study. First, ArcGIS spatial analysis techniques were used to extract the flood area; information gridding technology was adopted to grid the affected area; different classes of the damage area and submerged depth were classified and extracted. Secondly, field data was collected to build the affected sample database. Then, a rapid disaster damage assessment model based on spatial information grid and BP neural network was set up and realized through the use of program code. Finally, the assessment system based on model development was used to evaluate the economic loss of the flood in Duchang County. The results show that the evaluating results calculated by the model method are close to the statistical value of flood damage, proving the reliability of the rapid flood damage assessment system based on spatial information grid and BP neural network.