基于空间信息格网和BP神经网络的洪灾损失评估研究

发布时间：2018-01-11 16:02

本文关键词：基于空间信息格网和BP神经网络的洪灾损失评估研究　出处：《江西理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：洪涝灾害是我国最严重的自然灾害之一,它具有发生频次高、破坏性大以及影响范围广等特点。洪涝灾害给我国带来的人员伤亡和经济损失,已经严重地制约了我国社会经济的可持续发展。对洪灾经济损失的有效评估不仅能够对防洪减灾行为与防洪工程所发挥出的效益进行准确评价,而且还能为抗洪救灾提供重要的决策依据。但是当前洪灾损失评估方法存在计算量大、操作繁琐以及评估精度低等问题,为此开展了基于空间信息格网和BP神经网络的洪灾损失评估研究,本文的主要工作内容和创新点如下:(1)使用空间信息格网技术应用在洪灾损失评估中,首先将受灾区域划分成洪水特性格网和社会经济展布格网;然后使用叠加分析方法将两者进行叠加分析生成进行洪灾损失评估的空间信息格网;最后应用C#和Arc Engine实现了基于空间信息格网的洪灾损失评估系统。(2)使用BP神经网络技术应用在洪灾损失评估中,首先收集好洪灾损失评估原始数据;然后提取洪灾评估影响因子;其次构建BP神经网络,并在此基础上构建洪灾损失评估模型;最后使用编程手段实现了基于BP神经网络的洪灾损失评估系统。(3)利用空间信息格网在淹没水深提取以及淹没面积统计的特点,结合BP神经网络能够对洪灾样本数据逐一进行归一化、训练、测试并得出预测值的优势,建立基于空间信息格网与BP神经网络的洪灾损失评估模型。(4)使用Arc GIS Engine开发引擎基于C#编程语言,并在处理好样本数据的基础上实现基于空间信息格网和BP神经网络的洪灾损失评估系统。应用开发好的洪灾损失评估系统针对鄱阳湖区某县2013年的洪灾进行了洪灾损失评估,得出的评估结果与当年实际经济损失结果接近,误差率较小。
[Abstract]:Flood and waterlogging is one of the most serious natural disasters in China, which has the characteristics of high frequency, great destruction and wide range of influence. The flood and waterlogging disaster brings casualties and economic losses to our country. It has seriously restricted the sustainable development of China's social economy. The effective evaluation of flood economic losses can not only accurately evaluate the flood prevention and mitigation behavior and the benefits of flood control projects. It can also provide an important decision basis for flood control and disaster relief. However, the current flood loss assessment method has many problems, such as large amount of calculation, cumbersome operation and low evaluation accuracy. For this reason, the research on flood damage assessment based on spatial information grid and BP neural network is carried out. The main contents and innovations of this paper are as follows: 1) the use of spatial information grid technology in flood disaster loss assessment. First, the affected area is divided into flood characteristic grid and socio-economic grid; Then the superposition analysis method is used to generate the spatial information grid for flood damage assessment. Finally, the flood damage assessment system based on spatial information grid is implemented by using C # and Arc Engine. BP neural network technology is used in flood damage assessment. First, collect the original data of flood damage assessment; And then extract the impact factors of flood assessment; Secondly, BP neural network is constructed, and on this basis, flood loss assessment model is constructed. Finally, a flood loss assessment system based on BP neural network is realized by programming method. It uses spatial information grid to extract the depth of submerged water and to calculate the submergence area. The BP neural network can normalize, train, test and get the predictive value of flood data one by one. A flood damage assessment model based on spatial information grid and BP neural network is established. (4) Arc GIS Engine development engine is developed based on C # programming language. The flood disaster loss assessment system based on spatial information grid and BP neural network is realized on the basis of processing the sample data. The application of the developed flood damage assessment system is aimed at the flood disaster in 2013 in a county in Poyang Lake region. A flood damage assessment was carried out. The result of evaluation is close to that of actual economic loss, and the error rate is small.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P426.616;P208;TP18

【参考文献】