摘要
以H市供水管网为例,从GIS系统数据库中选取完整性较高的管材、管径、管长、管龄与管位等参数,运用XGBoost算法对管网进行健康评估,结果表明,模型准确率为86.0%、AUC值为0.93,评估效果良好;同时结合实际排查结果,得出管网修复指数公示为RI=0.775×10×F+0.15×K+0.075×E。
Based on the analysis of a city’s water distribution network database,while selecting parameters such as pipe diameter,length,age and pipe position with highly complete information from GIS database,XGBoost algorithm was used to evaluate the health of the pipeline.The well-being results outputed that accuracy of model was 86.0%and AUC was 0.9.Meanwhile,the pipe restoration factor formula RI=0.775×10×F+0.15×K+0.075×E was obtained combining actual investigation results.
作者
周华
姜石磊
周琮辉
应松枝
徐昕吟
翁晓丹
ZHOU Hua;JIANG Shilei;ZHOU Conghui;YING Songzhi;XU Xinyin;WENG Xiaodan(Huadong Engineering Co.,Ltd.,Power China,Hangzhou 311100,China;Hangzhou Water Group Co.,Ltd.,Hangzhou 310000,China;Hangzhou Water Design Institute Co.,Ltd.,Hangzhou 310000,China)
出处
《给水排水》
CSCD
北大核心
2021年第S02期491-494,共4页
Water & Wastewater Engineering
关键词
供水管网
风险评估
修复指数
XGBoost算法
管道维护
Water supply system
Risk assessment
Restoration factor
XGBoost algorithm
Pipeline maintenance
作者简介
通信作者:周华,男,1984年出生,浙江宁波人,硕士研究生,高级工程师。主要研究方向为供水系统规划,供水厂及供水管网的优化设计。
