[1]李旭杰,史灵,花思洋,等.基于随机森林的水质监测指标预测[J].江苏水利,2022,(05):6-10.
 LI Xujie,SHI Ling,HUA Siyang,et al.Water quality monitoring indicators prediction based on random forests[J].JIANGSU WATER RESOURCES,2022,(05):6-10.
点击复制

基于随机森林的水质监测指标预测()
分享到:

《江苏水利》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2022年05期
页码:
6-10
栏目:
水生态与环境
出版日期:
2022-06-01

文章信息/Info

Title:
Water quality monitoring indicators prediction based on random forests
文章编号:
1007-7839(2022)05-0006-0005
作者:
李旭杰12史灵23花思洋3孙颖24黄凤辰2
(1.河海大学 海洋与近海工程研究院,江苏 南通 226300;2.河海大学 计算机与信息学院,江苏 南京 210098;3.钛能科技股份有限公司,江苏 南京 211806;4.江苏开放大学 信息工程学院,江苏 南京 210017)
Author(s):
LI Xujie123 SHI Ling24 HUA Siyang4 SUN Ying25 HUANG Fengchen2
(1.Institute of Ocean and Offshore Engineering, Hohai University, Nantong 226300, China;2.College of Computer and Information, Hohai University, Nanjing 210098, China;3.Talent Science and Technology Co., Ltd., Nanjing 211806, China;4.School of Information Engineering, Jiangsu Open University, Nanjing 210017, China)
关键词:
Pearson相关系数多元线性回归算法随机森林模型秦淮新河
Keywords:
Pearson correlation coefficient multiple linear regression algorithm random forest model Qinhuai New River
分类号:
X522
文献标志码:
B
摘要:
通过采集2020年6月至2021年6月南京市秦淮新河代表站的DO、WT、pH、COD、NH3-N、TUR 6类水质监测指标数据,利用Pearson相关系数对监测指标间的相关程度进行分析,从而得到各监测指标间的相关系数,进一步通过多元线性回归算法得到高度相关的参数指标间的统计关系,利用回归方程的形式表示监测变量间的因果关系,最后通过随机森林算法利用水质监测中的自变量指标实现对因变量指标的预测,达到减少监测项目从而降低监测成本的目的。研究结果表明因变量水质监测指标的预测值和实际值几乎重合,有效说明随机森林模型能够实现因变量水质监测指标的准确预测。
Abstract:
Through collecting the data of six kinds of water quality monitoring indicators of Qinhuai New River representative station from June 2020 to June 2021, including dissolved oxygen, water temperature, PH value, chemical oxygen demand, ammonia nitrogen and turbidity. Pearson correlation coefficient was used to analyze the correlation degree among monitoring indicators, so as to obtain the correlation coefficient among monitoring indicators. Further through multiple linear regression algorithm was highly related to the statistical relationship among the parameters, using the regression equation in the form of said monitoring the causal relationship among variables, finally by random forest algorithm using water quality monitoring in the independent variable of the dependent variable indicators forecast, achieve the goal of reducing monitoring project so as to reduce the monitoring cost. The results show that the predicted value and the actual value of the dependent variable water quality monitoring index almost coincide, which effectively indicates that the random forest model can achieve the accurate prediction of the dependent variable water quality monitoring index.

参考文献/References:

[1]邵园园,戴庆云,蒋涛,等. 秦淮河流域水生态(环境)调度研究[J]. 江苏水利,2020(12):44-47,69.
[2]马振,周密. 聚类分析在秦淮河水质指标相关性研究中的应用[J].水文,2018,38(1):77-80.
[3]WEIQI H E. Water quality monitoring in a slightly-polluted inland water body through remote sensing-casestudy of the Guanting Reservoir in Beijing,China[J]. Frontiers of Environmental Science & Engineering in China,2008,2(2):163-171.
[4]郑德论. 汕头龙湖沟河段水质状况监测与评价对策[J].广州化工,2017,45(20):117-119.
[5]汤云. 闽江流域水质时空分布特征分析及水质预测[D]. 福州:福州大学,2018.
[6]杨娜,陈国鹰. 基于灰色聚类分析的白洋淀水质评价[J].科技创新与应用,2019(34):43-45.
[7]秦无双,胡艳丽,肖娟,等. 蓬溪县主要地表水体水质现 状及水质改善对策[J]. 绿色科技,2020(12):61-63.
[8]杨志民. 契爷石水库水质监测评价及水质预测研究[J].中国水能及电气化,2021(11):64-68.
[9]刘玲,孙玉豪. 基于Pearson & BIM云的工程造价信息 数据共享研究[J]. 工程经济,2021,31(8):20-24.
[10]贾俊平,何晓群,金勇进. 统计学[M].6版.北京:中国人民大学出版社,2014.
[11]彭一晋. 基于智能手机软件数据对用户性别的预测[D]. 大连:大连理工大学,2021.
[12]刘海军,刘韵锋,陈侨,等. 基于随机森林和贝叶斯优化的TiO_2光催化污染物降解速率预测模型研究[J]. 信息记录材料,2021,22(8):25-27.

备注/Memo

备注/Memo:
收稿日期:2021-12-28
基金项目:江苏省水利科技项目(2020028);南通市社会民生科技项目(MS22021042);广东省水利科技创新项目(2020-04);江苏省教育厅未来网络科研基金资助(FNSRFP-2021-YB-7);中国科学院无线传感网与通信重点实验室开放课题(20190914)
作者简介:李旭杰(1979—),男,副教授,博士,主要从事水利信息化技术研究。E-mail:lixujie@hhu.edu.cn
更新日期/Last Update: 2022-05-15