Construction of an Intelligent Risk Identification Model for Rural Water Supply Projects and Analysis of the Influencing Factors

ABSTRACT Enhancing risk identification and early warning capabilities for rural water supply projects is of great significance for promoting high‐quality development in rural water services. This study developed an intelligent risk assessment model using the XGBoost algorithm and SHAP interpretability framework based on data from Wuxi County, Chongqing, China. The results demonstrate high model accuracy (F1 = 0.89) and identify climate factors as primary risk sources, with daily rainfall (SHAP = 0.91) and maximum temperature (SHAP = 0.83) showing significant impacts. Pipeline service life (SHAP = 0.86) was found to substantially affect system reliability. The study reveals that climate change and infrastructure aging significantly influence rural water supply security. In comparison, factors related to water supply scale, water quality and payment methods showed secondary importance, while water source type, purification methods and operational management had minimal effects. The findings suggest promoting large‐scale centralized water supply systems and standardized upgrades for small‐scale water projects to enhance resilience against rural water supply risks.