基于SVM参数优化的饮用水水源地风险预警
作者:
作者单位:

(1.水利部珠江水利委员会珠江水利科学研究院,广东 广州 510611;2.华南理工大学土木与交通学院,广东 广州 510641;3.长江生态环保集团有限公司,湖北 武汉 430062;4.郑州大学水利与交通学院,河南 郑州 450001 )

作者简介:

唐红亮(1986—),男,高级工程师,博士研究生,主要从事水环境治理及水资源保护与开发利用研究。E-mail:464373552@qq.com通信作者:王冬(1998—),男,硕士,主要从事水资源开发利用与管理研究。E-mail:1397032623@qq.com

基金项目:

国家科技基础资源调查专项(2019FY101900);国家重点研发计划项目(2022YFC3202200, 2021YFC3200205 )


Risk warning of drinking water source areas based on SVM parameter optimization//
Author:
Affiliation:

(1.Pearl River Water Resourcs Research Institute, Pearl River Water Resources Commission of the Ministry of Water Resources, Guangzhou 510611, China;2.School of Civil Engineering & Transportation, South China University of Technology, Guangzhou 510641, China;3.Yangtze Ecology and Environment Co., Ltd., Wuhan 430062, China;4.School of Water Conservancy and Transportation, Zhengzhou University, Zhengzhou 450001, China)

  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 文章评论
    摘要:

    以茂名市电白区罗坑水库水源地为例,运用随机森林模型识别饮用水水源地风险预警指标,基于支持向量机(SVM)模型、遗传算法(GA)和蝴蝶优化算法(BOA)构建了3种饮用水水源地风险预警模型,并优选最佳模型对罗坑水库水源地风险状态进行了预警分析。结果表明:相较于SVM模型和GA-SVM模型,BOA-SVM模型的平均相对误差、均方误差、模型运行时间和均方根误差均有不同程度的下降,R2分别上升了5.16%和3.01%,BOA-SVM模型更适用于饮用水水源地风险预警;2023—2025年罗坑水库水源地风险预警级别均为轻警(蓝灯),处于较低风险状态;虽然风险预警级别相同,但风险预警指数逐年降低,2025年的风险预警指数为1.61,已逼近低风险状态的临界值。相关管理部门应加大现有规划、方案及政策的实施力度,进一步降低罗坑水库水源地风险等级。

    Abstract:

    Taking the water source area of Luokeng Reservoir in Dianbai District, Maoming City as an example, the random forest model was used to identify the risk warning indicators of drinking water sources. Three risk warning models for drinking water sources were constructed based on the SVM model, GA and BOA. The optimal model was selected to warn the risk status of Luokeng Reservoir water source area. The results show that compared with the SVM model and the GA-SVM model, the mean relative error, mean square error, model operation time, and root mean square error of the BOA-SVM model have all decreased to varying degrees. The R2 values have increased by 5.16% and 3.01% respectively. The BOA-SVM model is more suitable for the risk warning of drinking water sources.From 2023 to 2025, the risk warning levels of the water source area of Luokeng Reservoir are all at the light warning level

    参考文献
    相似文献
    引证文献
引用本文

唐红亮,王冬,董延军,等.基于SVM参数优化的饮用水水源地风险预警[J].水资源保护,2025,41(2):209-215, 232.(TANG Hongliang, WANG Dong, DONG Yanjun, et al. Risk warning of drinking water source areas based on SVM parameter optimization//[J]. Water Resources Protection,2025,41(2):209-215, 232.(in Chinese))

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-05-30
  • 在线发布日期: 2025-04-14