LIGengyan,ZHAOMengke,LIUHaixing.Leakage Assessment Model for Water Supply Networks Based on Pipe Segment Grouping[J].China Water & Wastewater,2026,42(5):63-68.
Leakage Assessment Model for Water Supply Networks Based on Pipe Segment Grouping
China Water & Wastewater[ISSN:1000-4062/CN:12-1073/TU]
volume:
第42卷
Number:
第5期
Page:
63-68
Column:
Date of publication:
2026-03-01
- Keywords:
- water supply network; pipe segment grouping; light gradient boosting machine algorithm; leakage assessment; Shapley additive explanation
- Abstract:
- To address the issue of poor data quality in the fundamental dataset of water supply networks, a pipe segment grouping-based method for leakage assessment was proposed. This methodology groups pipes based on their geographic location and intrinsic characteristics. Based on the grouping results, a pipe-group dataset was constructed. Using the light gradient boosting machine (LightGBM) algorithm, a leakage assessment model for pipe groups was developed and applied to a case study of a municipal water supply network. The results showed that the constructed model demonstrated excellent performance, with an area under the receiver operating characteristic curve (AUC) exceeding 0.86. Compared to the pipe segment-based leakage assessment models found in existing studies, the proposed pipe group model demonstrates significantly superior performance, achieving an AUC increase of 0.13. Shapley additive explanation (SHAP) analysis revealed that pipe length, leakage time interval, and historical leakage frequency were the more important influencing factors in the assessment model.
Last Update:
2026-03-01