基于机器学习的人工湿地出水水质预测与影响因素

陈亚松; 刘家雯; 赵云鹏; 周英萍; 沈秋实; 肖琳; 钱新

PDF(2179 KB)

中国环境科学 ›› 2025, Vol. 45 ›› Issue (6) : 3161-3170.

水污染与控制

基于机器学习的人工湿地出水水质预测与影响因素

陈亚松¹, 刘家雯², 赵云鹏¹, 周英萍², 沈秋实¹, 肖琳², 钱新²

作者信息 +

Prediction of effluent water quality and analysis of influencing factors in constructed wetlands based on machine learning

CHEN Ya-song¹, LIU Jia-wen², ZHAO Yun-peng¹, ZHOU Ying-ping², SHEN Qiu-shi¹, XIAO Lin², QIAN Xin²

Author information +

文章历史 +

摘要

基于水质指标、气候指标、湿地运行参数3个方向,收集以往研究文献数据,通过3种机器学习模型预测人工湿地出水氨氮(NH₄⁺-N)、COD、磺胺甲噁唑(SMX)以及部分重金属的浓度.结果表明,随机森林(Random Forest)在整体性能上略优于XGBoost和LightGBM,其决定系数(R²)和均方根误差(RMSE)的表现更为稳定,尤其是在NH₄⁺-N和SMX的预测上取得更高精度(NH₄⁺-N预测的R²分别为0.93、0.89和0.87).相比之下,在COD的预测中,3种模型的表现相对较弱,R²分别为0.71、0.61、0.64.通过引入SMOTE数据扩充技术,模型的预测性能和精度得到了显著的提升,尤其是对COD的预测性能提升幅度达7.04%~26.23%.本研究将数据分析与机器学习算法相结合,可为实际工程应用提供可行方法.

Abstract

Based on water quality indicators, climate indicators, and wetland operation parameters, data from previous studies were collected to predict the effluent concentrations of ammonia nitrogen (NH₄⁺-N), COD, sulfamethoxazole (SMX), and some heavy metals in constructed wetlands using three machine learning models. The results showed that the Random Forest model slightly outperformed XGBoost and LightGBM in overall performance, demonstrating more stable R² and RMSE values. In particular, it achieved higher accuracy in predicting NH₄⁺-N and SMX concentrations, with R² values of 0.93, 0.89, and 0.87, respectively, for NH₄⁺-N. In contrast, the models performed relatively weaker in COD predictions, with R² values of 0.71, 0.61, and 0.64, respectively. By incorporating the SMOTE data augmentation technique, the prediction performance and accuracy of the models were significantly enhanced, especially for COD, where improvements ranged from 7.04% to 26.23%. This study combines scientific data analysis with machine learning algorithms, providing a feasible approach for practical engineering applications.

导出引用

陈亚松, 刘家雯, 赵云鹏, 周英萍, 沈秋实, 肖琳, 钱新. 基于机器学习的人工湿地出水水质预测与影响因素[J]. 中国环境科学. 2025, 45(6): 3161-3170

CHEN Ya-song, LIU Jia-wen, ZHAO Yun-peng, ZHOU Ying-ping, SHEN Qiu-shi, XIAO Lin, QIAN Xin. Prediction of effluent water quality and analysis of influencing factors in constructed wetlands based on machine learning[J]. China Environmental Science. 2025, 45(6): 3161-3170

中图分类号： X703

参考文献

[1] Nguyen X C, Chang S W, Nguyen T L, et al. A hybrid constructed wetland for organic-material and nutrient removal from sewage: Process performance and multi-kinetic models [J]. Journal of environmental management, 2018,222:378-384.
[2] Saeed T, Sun G. The removal of nitrogen and organics in vertical flow wetland reactors: predictive models [J]. Bioresource Technology, 2011, 102(2):1205-1213.
[3] Langergraber G, Rousseau D P, Garcia J, et al. CWM1: a general model to describe biokinetic processes in subsurface flow constructed wetlands [J]. Water Science and Technology, 2009,59(9):1687-1697.
[4] Nguyen X C, Nguyen T P, Lam V S, et al. Estimating ammonium changes in pilot and full-scale constructed wetlands using kinetic model, linear regression, and machine learning [J]. Science of the Total Environment, 2024,907, doi:10.1016/j.scitotenv.2023.168142.
[5] 董越洋,徐波,王鹏,等.一种基于Stella和R语言的湿地氮素动力学模型[J]. 中国环境科学, 2020,40(1):198-205.Dong Y Y, Xu B, Wang P, et al. An ecological kinetic model of nitrogen cycle in the wetland based on Stella and R language [J]. China Environmental Science, 2020,40(1):198-205.
[6] Yang B, Xiao Z, Meng Q, et al. Deep learning-based prediction of effluent quality of a constructed wetland [J]. Environmental Science and Ecotechnology, 2023,13, doi:10.1016/j.ese.2022.100207.
[7] Vu H L, Ng K T W, RichterA, et al. Analysis of input set characteristics and variances on k-fold cross validation for a Recurrent Neural Network model on waste disposal rate estimation [J]. Journal of Environmental Management, 2022,311,doi:10.1016/j.jenvman. 2022.114869.
[8] Wang Z, Wu F, Hao N, et al. The combined machine learning model SMOTER-GA-RF for methane yield prediction during anaerobic digestion of straw lignocellulose based on random forest regression [J]. Journal of Cleaner Production, 2024,466,doi:10.1016/j.jclepro.2024. 142909.
[9] Marcilio W E, Eler D M. From explanations to feature selection: assessing SHAP values as feature selection mechanism [Z]. 202033rd SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI). 2020,340(7),doi:10.1109/sibgrapi51738.2020.00053.
[10] 杨晓玲,王梦晓,李晓娟,等.机器学习-多元线性回归预测铜的水生态基准[J]. 中国环境科学, 2024,44(7):3976-3985.Yang X L, Wang M X, Li X J, et al. Predicting the water ecological criteria of copper using machine learning and multiple linear regression approach [J]. China Environmental Science, 2024,44(7): 3976-3985.
[11] 周琪,于洋,刘苗苗,等.基于机器学习和非参数估计的PM_2.5风险评估[J]. 中国环境科学, 2022,42(8):3554-3560.Zhou Q, Yu Y, Liu M M, et al. Risk assessment of PM_2.5 pollution based on machine learning and nonparametric estimat [J]. China Environmental Science, 2022,42(8):3553-3560.
[12] 李涛宇,许秀春,杨轩,等.利用机器学习预测华北地区冬小麦农田氮淋失[J]. 中国环境科学, 2025,45(1):343-354.Li T Y, Xu X C, Yang X, et al. Prediction of nitrogen leaching from winter-wheat production in North China based on random forest and XGBoost [J]. China Environmental Science, 2025,45(1):343-354.
[13] Li H, Xie L, Zhou B, et al. Machine learning-assisted prediction and identification of key factors affecting nitrogen metabolism for aerobic granular sludge [J]. Environmental Research, 2025,273,doi:10.1016/ j.envres.2025.121158.
[14] Bao Y, Wang Y, Hu M, et al. Deciphering the impact of cascade reservoirs on nitrogen transport and nitrate transformation: Insights from multiple isotope analysis and machine learning [J]. Water Research, 2025,268, doi:10.1016/j.watres.2024.122638.
[15] Cosenza A, Mannina G, Vanrolleghem P A, et al. Global sensitivity analysis in wastewater applications: A comprehensive comparison of different methods [J]. Environmental Modelling and Software, 2013, 49:40-52.
[16] 苏子龙,严文亮,李慧敏,等.基于改进麻雀搜索算法优化BP神经网络的农业碳排放预测[J]. 环境科学:1-15.Su Z L, Yan W L, Li H M, et al. Prediction of Agricultural Carbon Emission Based on Improved BP Neural Networkwith Optimized Sparrow Search Algorith [J]. China Environmental Science: 1-15.
[17] Singh S, Kulshreshtha N M, Goyal S, et al. Performance prediction of horizontal flow constructed wetlands by employing machine learning [J]. Journal of Water Process Engineering, 2022,50,doi:10.1016/j.jwpe. 2022.103264.
[18] Rampuria A, Gupta A B, Brighu U. Nitrogen transformation processes and mass balance in deep constructed wetlands treating sewage, exploring the anammox contribution [J]. Bioresource Technology, 2020,314,doi:10.1016/j.biortech.2020.123737.
[19] Akratos C S, Papaspyros J N E, Tsihrintzis V A. Total nitrogen and ammonia removal prediction in horizontal subsurface flow constructed wetlands: Use of artificial neural networks and development of a design equation [J]. Bioresource Technology, 2009,100(2):586-596.
[20] Salem M, El-sayed Gabr M, Mossad M, et al. Random Forest modelling and evaluation of the performance of a full-scale subsurface constructed wetland plant in Egypt [J]. Ain Shams Engineering Journal, 2022,13(6),doi:10.1016/j.asej.2022.101778.
[21] Dai W, Pang J-W, Zhao Y-j, et al. Machine learning assisted combined systems of wastewater treatment plants with constructed wetlands optimal decision-making [J]. Bioresource Technology, 2024,399, doi:10.1016/j.biortech.2024.130643.
[22] Fernandez A, Garcia S, Herrera F, et al. SMOTE for learning from imbalanced data: progress and challenges, marking the 15-year anniversary [J]. Journal of artificial intelligence research, 2018,61: 863-905.
[23] 何强,胡书山,向泽毅,等.垂直流人工湿地系统净化污水厂尾水脱氮效果[J]. 中国环境科学, 2023,43(8):3956-3965.He Q, Hu S S, Xiang Z Y, et al. Study on the nitrogen removal ability of vertical flow constructed wetland treating tailwater of sewage plant [J]. China Environmental Science, 2023,43(8):3956- 3965.
[24] 王旭,郭雯,王明果,等.人类活动影响下异龙湖浮游植物碳氮稳定同位素变化特征[J]. 中国环境科学, 2023,43(6):3087-3099.Wang X, Guo W, Wang M G, et al. Pattern in changes of carbon and nitrogen stable isotopes of phytoplankton in Yilong Lake under the influence of human activities [J]. China Environmental Science, 2023, 43(6):3087-3099.
[25] Zhou L, Xiang X, Chen Y, et al. Enhanced nitrogen removal in modular moving bed constructed wetland at low temperature: Optimization of dissolved oxygen distribution and reconfiguration of core microbial symbiosis [J]. Environmental Research, 2025,276, doi:10.1016/j.envres.2025.121507.
[26] 刘冰,郑煜铭,秦会安,等.填料对潮汐流人工湿地中CANON作用强化的影响[J]. 环境科学, 2021,42(1):283-292.Liu Bing, Zhen Y M, Qin H A, et al. Effect of filter medium on the enhancement of complete autotrophic nitrogen removal over nitrite process in a tidal flow constructed wetland [J]. Environmental Science, 2021,42(1):283-292.
[27] Wang X, Shen Z, Zhang Q, et al. Critical role of biochar in the production and emission of greenhouse gas N₂O in constructed wetlands: A comprehensive review [J]. Journal of Cleaner Production, 2025,505,doi:10.1016/j.jclepro.2025.145487.
[28] Ouyang B, Zhang Z, Chen F, et al. Energy production and denitrogenation performance by sludge biochar based constructed wetlands-microbial fuel cells system: Overcoming carbon constraints in water [J]. Water Research, 2025,273, doi:10.1016/j.watres.2024. 123024.
[29] 苏建成.地表水中温度、溶解氧、氨氮的关系研究[J]. 科技创新与应用, 2020,(14):44-55.Su J C. Research on the relationship between temperature, dissolved oxygen, and ammonia nitrogen in surface water [J]. Technology Innovation and Application, 2020,(14):44-55.
[30] Feng L, He S, Yu H, et al. A novel plant-girdling study in constructed wetland microcosms: Insight into the role of plants in oxygen and greenhouse gas transport [J]. Chemical Engineering Journal, 2022,431, doi:10.1016/j.cej.2021.133911.
[31] Majumder A, Bhatnagar A, Kumar Gupta A. Simultaneous removal of sulfamethoxazole, 17β-estradiol, and carbamazepine from hospital wastewater using a combination of a continuous constructed wetland- based system followed by photocatalytic reactor [J]. Chemical Engineering Journal, 2023,466,doi:10.1016/j.cej.2023.143255.
[32] Gulkowska A, Leung H W, So M K, et al. Removal of antibiotics from wastewater by sewage treatment facilities in Hong Kong and Shenzhen, China [J]. Water Research, 2008,42(1/2):395-403.
[33] Chen H, Ailijiang N, Cui Y, et al. Enhanced removal of PPCPs and antibiotic resistance genes in saline wastewater using a bioelectrochemical-constructed wetland system [J]. Environmental Research, 2024,260, doi:10.1016/j.envres.2024.119794.
[34] Yu K, Hei S, Li P, et al. Removal of intracellular and extracellular antibiotic resistance genes and virulence factor genes using electricity- intensified constructed wetlands [J]. Journal of Hazardous Materials, 2024,475,doi:10.1016/j.jhazmat.2024.134749
[35] Liu Y, Liu X, Wang H, et al. Performance and mechanism of SMX removal in an electrolysis-integrated tidal flow constructed wetland at low temperature [J]. Chemical Engineering Journal, 2022,434, doi:10.1016/j.cej.2022.134494.