基于管材分类的供水管网爆管风险预测模型  

Model of Risk Prediction for Water Supply Pipelines Burst Based on Pipe Material Classification

作　　者：陈能 信昆仑[1] 莫宇琨 

机构地区：[1]同济大学环境科学与工程学院,上海200092

出　　处：《净水技术》2018年第1期112-116,共5页Water Purification Technology

基　　金：国家重点研发计划项目(2016YFC0400602);国家自然科学基金项目(51678425)

摘　　要：以ZZ市的供水管网数据为例,根据ZZ市管道爆管历史记录的数据特征,采用了按管材分组的方式,并结合经典的Cox比例风险模型为每组分别建立爆管风险预测模型。将此模型与ZZ市GIS管网进行结合,对该市管网进行爆管风险评估,得到了管道爆管风险分布图,并在此基础上结合历史爆管点位的空间分布对模型的预测结果进行验证。结果表明,模型对高危险和危险管道的预测准确率接近70%,说明该模型能有效地识别出管网中爆管风险较高的管道。According to water supply networks and characteristics of pipe burst historic record of ZZ City, the prediction model of pipe burst risk is set up for each group by pipe materials classification combining with the classic Cox proportional hazard model. Then the model, combining with GIS network of ZZ City, is used for obtaining the distribution of pipe burst risk to assess the pipe burst risk and is validated with the spatial distribution of historic pipe burst points. The results show that the prediction accuracy of the model for high risk and dangerous pipe is close to 70%, which shows that the model can effectively identify the pipes with higher pipe burst risk in the pipe network.

关 键 词：给水管网 爆管 风险预测模型 COX比例风险模型 

分 类 号：TU991.1[建筑科学—市政工程]