基于机器学习的城市污水管道缺陷检测效率优化框架研究  

Research on optimization framework for urban sewage pipeline defect detection efficiency based on machine learning

作　　者：马辉 贺鹰霞 MA Hui;HE Yingxia(School of Economics and Management,Tianjin Urban Construction University,Tianjin 300384,China)

机构地区：[1]天津城建大学经济与管理学院,天津300384

出　　处：《工业安全与环保》2025年第4期53-58,63,共7页Industrial Safety and Environmental Protection

基　　金：天津市哲学社会科学规划课题(TJGL24-016);天津市研究生科研创新项目(2022SKYZ330)。

摘　　要：为了解决传统污水管道缺陷检测流程存在的效率低下、资源浪费以及高强度工作下人工漏检率高等问题,使用机器学习算法构建了一套针对地下污水管网缺陷的两阶段诊断方法。首先利用管道历史数据构建缺陷发生概率估算模型,识别高危管段,而后针对高危管道实行进一步的管道潜在缺陷智能检测。以广东省中山市某区域为实例,验证模型对高危管段识别、潜在缺陷诊断的精确性。结果表明,构建的城市地下污水管网缺陷隐患两阶段排查方法有效提高了地下污水管道缺陷的整体排查效率。A two-stage diagnostic method for underground sewage pipeline defects was developed using machine learning algorithms to address the inefficiency,resource wastage,and high human error rates in traditional defect detection processes.First,a probability estimation model was constructed using historical pipeline data to identify highrisk pipeline sections.Then,intelligent detection was applied to identify potential defects in these high-risk sections.A case study in Zhongshan City,Guangdong Province,China,validated the accuracy of the model in identifying highrisk sections and diagnosing potential defects.The results show that the two-stage inspection method effectively improve the overall efficiency of underground sewage pipeline defect detection.

关 键 词：污水管道 缺陷检测 机器学习 效率优化 

分 类 号：TU9[建筑科学]