随机森林模型在城市排水分区排水能力评估中的应用研究

Research on the Application of Random Forest Model in Evaluating the Drainage Capacity of Urban Drainage Zoning

我国城市内涝灾害频发,城市排水防涝能力建设是内涝治理工作的重要环节,评估城市排水能力,识别薄弱环节,对分时序分重点开展内涝治理工作至关重要。本文以福州市江北城区为例,以排水分区为单元,利用随机森林模型评估其排水能力。结果表明:随机森林模型评估精度达到90%,AUC值为0.883,模型具有较高的评估精度,评估结果能够较好地反映实际积水情况。

Urban water-logging disasters occur frequently in China.The construction of urban drainage and water-logging prevention capacity is an important part of water-logging control.It is very important to evaluate urban drainage capacity and identify weak links in order to carry out water-logging control work in time sequence and focus.Taking Jiangbei urban area of Fuzhou as an example,taking drainage zoning as a unit,this paper uses random forest model to evaluate its drainage capacity.The results show that:the evaluation accuracy of random forest model is 90%,the AUC value is 0.883,the model has high evaluation accuracy,and the model evaluation results can better reflect the actual ponding situation.