[1]刘 飞,凌洪军,路广宽.基于改进随机森林算法的通榆河水生态健康状况评价研究[J].江苏水利,2019,(10):18-22+68.
　LIU Fei,LING Hongjun,LU Guangkuan.Study on the evaluation of water ecological health status of Tongyu River based on improved random forest algorithm[J].JIANGSU WATER RESOURCES,2019,(10):18-22+68.
点击复制

基于改进随机森林算法的通榆河水生态健康状况评价研究()
分享到：

《江苏水利》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2019年10期
页码:
18-22+68
栏目:
水生态与环境
出版日期:
2019-11-02

文章信息/Info

Title:
Study on the evaluation of water ecological health status of Tongyu River based on improved random forest algorithm
文章编号:
1007-7839(2019)10-0018-05
作者:
刘 飞1;  凌洪军2;  路广宽3
1.江苏省建湖县沿河水利管理服务站, 江苏 盐城 224700; 2.江苏省建湖县水利局, 江苏 盐城 224700; 3.江苏省建湖县通榆河管理所, 江苏 盐城 224700
Author(s):
LIU Fei1;  LING Hongjun2;  LU Guangkuan3
1.Yanhe Water Conservancy Management Service Station of Jianhu County, Yancheng 224700, Jiangsu; 2.Jianhu Water Conservancy Bureau, Yancheng 224700, Jiangsu; 3.Tongyu River Management Office of Jianhu County, Yancheng 224700, Jiangsu
关键词:
改进随机森林算法;  水生态健康;  评价体系;  通榆河
Keywords:
improved random forest algorithm(IBF);  water ecology health;  evaluation system;  Tongyu River
分类号:
X826
文献标志码:
B
摘要:
通过改进传统的随机森林算法,对江苏省通榆河的水生态健康状况进行评价研究,建立了包含18个具体指标的通榆河水生态健康状况评价体系,并将评价结果分为“病态、微病态、亚健康、微健康、健康”5个等级。模型改进结果表明,改进的随机森林算法(IRF)较传统的随机森林(RF)算法和径向基人工神经网络(ANN-RBF)在模型误差和计算效率上均有明显提升,算法性能得到较大改善; 评价结果显示,通榆河在2016～2018年的水生态健康状况分别为微病态、亚健康、微健康,整体呈现良好的发展态势。改进的模型可为相关评价研究提供借鉴和指导。
Abstract:
The water ecological health status of Tongyu River in Jiangsu Province was evaluated by improved random forest algorithm, the evaluation system of Tongyu River water ecological health status including 18 specific indicators was established, and the evaluation results were divided into 5 grades of morbidity, micro-morbidity, sub-health, micro-health, and health.The improved model results showed that IRF had obvious improvement in model error and computational efficiency compared with RF and ANN-RBF, and the algorithm performance was greatly improved.The evaluation results showed that the water ecological health status of Tongyu River in 2016-2018 was micro-pathological, sub-health and micro-health, and the overall development trend was good.The improved model could provide reference and guidance for relevant evaluation research.

参考文献/References:

[1] 刘萍.建湖县河湖管理范围划定工作研究[J].江苏水利, 2018(11):53-57.
[2] 镇云, 李友春.城市河道生态护岸稳定分析及应用[J].江苏水利, 2018(06):66-69+72.
[3] 鲁亚会.基于随机森林特征选择的贝叶斯分类模型及应用[D].郑州:华北水利水电大学, 2017.
[4] 余胜男, 陈元芳, 顾圣华, 等.随机森林在降水量长期预报中的应用[J].南水北调与水利科技, 2016, 14(01):78-83.
[5] 赖成光, 陈晓宏, 赵仕威, 等.基于随机森林的洪灾风险评价模型及其应用[J].水利学报, 2015, 46(01):58-66.
[6] 单红喜.基于模糊数学的河道健康综合评价方法研究[J].江苏水利, 2018(05):57-62.

备注/Memo

备注/Memo:
收稿日期:2019-06-17 作者简介:刘飞(1989—), 男,本科,主要从事河道治理和评价研究等相关工作。
更新日期/Last Update: 2019-11-15