排洪工程质量与安全协同监督路径探析

Analysis of collaborative supervision paths for quality and safety of flood discharge projects

摘要:

为提升中小型排洪工程质量与安全监督的科学性与智能化水平,解决传统人工巡检量化不足、协同性弱的问题,以贵州省瓮安县岩孔排洪工程为研究对象,通过层次分析法(AHP)确定权重,利用模糊综合评价(Fuzzy)量化不确定性,并利用深度神经网络(DNN)构建非线性决策模块,构建AHP–Fuzzy–DNN质量与安全协同监督模型。模型选取质量、安全、信息化与协同4个一级指标,基于124组监督样本完成模型训练。结果表明,协同监督模型对工程风险等级的预测准确率达92.6%,决定系数R²为0.926,排洪工程质量与安全协同监督综合评价指数均值为0.873,处于优良状态。研究可实现排洪工程监管的智能化、定量化与可视化,为水利工程数字化监管提供可行路径。

Abstract:

To enhance the scientific and intelligent level of quality and safety supervision for small- and medium-sized flood discharge projects and to address the limitations of traditional manual inspection, such as insufficient quantification and weak coordination, this study took the yankong flood discharge project in Weng’an County, Guizhou Province, as the research object. A collaborative supervision model for quality and safety was constructed using the Analytic Hierarchy Process (AHP), Fuzzy Comprehensive Evaluation, and a Deep Neural Network (DNN). AHP was employed for weight determination, fuzzy logic for quantifying uncertainties, and DNN for constructing a nonlinear decision-making module. The model incorporated four primary indicators—quality, safety, informatization, and coordination—and was trained using 124 sets of supervision samples. The results demonstrated that the model achieved a prediction accuracy of 92.6% for project risk levels, with a coefficient of determination (R²) of 0.926. The comprehensive evaluation index of quality and safty collaborative supervision for the flood discharge project was 0.873, indicating an excellent operational status. This research enables intelligent, quantitative, and visualized supervision of flood discharge projects, providing a feasible path for the digital transformation of hydraulic engineering regulation.