数据驱动的供水管网安全评估模型应用实践  

Application practice of data-driven safety assessment model of water supply pipeline network

作　　者：王云亮 乔庆 韩冰 WANG Yunliang;QIAO Qing;HAN Bing(Shanghai Chengtou Water Group Co.,Ltd.,Shanghai 200002,China;Sino-Italian Center for Sustainability of Tongji University,Shanghai 200092,China)

机构地区：[1]上海城投水务(集团)有限公司供水分公司,上海200002 [2]同济大学中意可持续发展中心,上海200092

出　　处：《给水排水》2023年第7期140-147,共8页Water & Wastewater Engineering

基　　金：上海城投水务(集团)有限公司基金项目(KY.GS.20.0002)。

摘　　要：以某示范区供水管网为对象,通过多源数据的收集与分析,采用数据驱动的机器学习模型实现供水管网的安全评估,确定不同管道的风险等级,为后续管道巡检和运维提供决策支持,为推进人工智能技术在“智慧水务”中的应用场景提供典型案例。A water supply pipeline network is taken as the demonstration project.Based on the collection and analysis of the multi-source data,a data-driven machine learning model is adopted for safety assessment of the pipeline network,which determines the risk level of different pipelines and hence provide decision support for the subsequent pipeline inspection and operation.It also presents a typical case to promote the application of artificial intelligence technology in smart water utilities.

关 键 词：供水管网 多源数据 安全评估 机器学习模型 

分 类 号：TU990[建筑科学—市政工程]