基于多源数据的城市洪涝灾害风险评估

Urban Flooding Risk Assessment Based on Multi-source Data

为定量分析城市洪涝灾害风险,克服传统城市洪涝灾害风险评估数据收集困难、评价指标定权难等问题,该文改进了一种指标体系评估法,从城市开源数据集和互联网中收集洪涝指标评估数据构建洪涝灾害风险评估模型,拓宽了数据采集的途径。首先基于随机森林算法对收集整理的城市数据指标进行特征重要性分析,完善了指标权重评估的定量分析;其次设计基于AHP层次分析法和熵权法计算城市洪涝指标的权重值使得权重值更加合理;最后结合洪涝数据与指标权重值对城市的洪涝灾害风险等级进行评估,并以南昌市为例进行模型的应用。结果表明洪涝灾害风险评估结果与实际洪涝灾害分布情况基本吻合,能够为洪涝灾害相关部门对洪涝灾害对预防与洪涝灾害研究提供理论依据、数据参考和数据支持。

In order to quantitatively analyze urban flood risk,overcome the traditional problems of difficult data collection and difficult weighting of urban flood risk assessment indicators,the paper improves an indicator system assessment method,collects flood indicator assessment data from urban open source data sets and the Internet,and builds a flood disaster risk assessment model,which broadens the way of data collection.Firstly,based on the random forest algorithm,the feature importance analysis of the collected urban data indicators is carried out,and the quantitative analysis of the indicator weight evaluation is improved.Secondly,the design is based on the AHP hierarchical analysis method and entropy weight method is used to calculate the weight value of the urban flood indicators,which makes the weight value more reasonable.Finally,the flood risk level of the city is evaluated by combining the flood data and the index weight value,and the model is applied by taking Nanchang as an example.The results show that the flood risk assessment results are basically consistent with the actual flooding distribution,and can provide theoretical basis,data reference and data support for flooding related departments on flooding prevention and flooding research.