JinXian,ZhangLei,MaJie,et al.Study on Synergistic Dosing of Dual Chemicals in Water Treatment Plants Based on MIMO Model[J].China Water & Wastewater,2026,42(15):61-66.
基于MIMO模型的供水厂双药协同投加研究
- Title:
- Study on Synergistic Dosing of Dual Chemicals in Water Treatment Plants Based on MIMO Model
- 关键词:
- 供水厂; 机器学习; 混凝剂投加量; 三氯化铁（FeCl3）; 聚合氯化铝（PAC）
- Keywords:
- water treatment plant; machine learning; coagulant dosage; ferric chloride (FeCl3); polyaluminum chloride (PAC)
- 摘要:
- 针对供水厂双药协同投加中难以同步精准确定投加量的问题，构建了基于多输入多输出（MIMO）架构的投加量预测模型。并以北方某大型供水厂为研究对象，采集2024年全年65 880组工艺运行与水质参数数据。模型采用XGBoost、SVR和RF-Elman三种算法，并引入多头注意力机制与惩罚项，以强化模型对pH、浊度等关键指标的响应能力；同时，采用平均绝对误差（MAE）、均方根误差（RMSE）和决定系数（R2）作为模型评价指标。结果显示，XGBoost在三氯化铁和聚合氯化铝投加量预测中表现最佳，测试集的R2分别达到0.992 4和0.982 8，RMSE与MAE均低于其他模型。特征重要性分析表明，进水温度、次氯酸钠与臭氧预氧化参数及构建的pH调节方程是关键影响因素。鲁棒性验证表明，XGBoost在极端水质条件下具有更低的水质超标风险与更高的工程安全性，综合表现最优。
- Abstract:
- A multi-input multi-output (MIMO)-based dosage prediction model was established to tackle the challenge of synchronous and accurate dosage determination in dual-chemical coordinated dosing for water treatment plants. A large-scale water treatment plant in northern China was taken as the research object, and 65 880 sets of operational and water quality data were collected over the entire year of 2024. Three algorithms—XGBoost, SVR, and RF-Elman—were adopted, and a multi-head attention mechanism along with a penalty term was introduced to strengthen the model’s responsiveness to key indicators such as pH and turbidity. Meanwhile, mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2) were employed as evaluation metrics. The results showed that XGBoost achieved the best predictive performance for both ferric chloride and polyaluminum chloride dosages, with test-set R2 values reaching 0.992 4 and 0.982 8, respectively, and RMSE and MAE both lower than those of the other models. Feature importance analysis indicated that influent temperature, sodium hypochlorite and ozone pre-oxidation parameters, and the constructed pH regulation equation were key influencing factors. Robustness validation demonstrated that XGBoost offered a lower risk of water quality exceedance and higher engineering safety under extreme water quality conditions, achieving the best overall performance.
相似文献/References:
[1]张杰,邱文心,陈超,等.新冠疫情控制期间供水厂消毒效果影响因素及其控制范围[J].中国给水排水,2021,37(24):1.
ZHANG Jie,QIU Wen-xin,CHEN Chao,et al.Discussion on Influencing Factors and Control Range of Disinfection Effect in Waterworks during the COVID-19 Outbreak[J].China Water & Wastewater,2021,37(15):1.
[2]林峰,李旭,曾翰,等.某大型城市智能供水调度优化算法研究与模拟实践[J].中国给水排水,2023,39(9):109.
LINFeng,LIXu,ZENGHan,et al.Optimized Algorithm of Intelligent Water Supply Scheduling and Its Simulation Practice in a Large City[J].China Water & Wastewater,2023,39(15):109.
[3]范鹏辉,姜涛,牛超群,等.基于液位监测及CNN-SVM的排水管网缺陷诊断[J].中国给水排水,2023,39(23):30.
FANPeng-hui,JIANGTao,NIUChao-qun,et al.Fault Diagnosis Method of Drainage Network Based on Liquid Level Monitoring Data and CNN-SVM[J].China Water & Wastewater,2023,39(15):30.
[4]刘霖皓,程麒铭,陈垚,等.生物滞留系统去除典型重金属的影响因素识别[J].中国给水排水,2023,39(23):124.
LIULin-hao,CHENGQi-ming,CHENYao,et al.Identification of Influencing Factors for Removal of Typical Heavy Metals in Bioretention System[J].China Water & Wastewater,2023,39(15):124.
[5]朱静怡,黄苑曦,衣启航,等.基于机器学习的高锰酸钾与有机污染物反应活性预测[J].中国给水排水,2024,40(11):41.
ZHUJing-yi,HUANGYuan-xi,YIQi-hang,et al.Prediction of Reactivity between Potassium Permanganate and Organic Pollutants Based on Machine Learning[J].China Water & Wastewater,2024,40(15):41.
[6]张静,田雨,齐天天,等.活性炭滤池对三氯甲烷的吸附、解吸与微生物作用[J].中国给水排水,2024,40(13):43.
ZHANGJing,TIANYu,QITian-tian,et al.Activated Carbon Filter for Adsorption and Desorption of Trichloromethane and Its Microbial Effects[J].China Water & Wastewater,2024,40(15):43.
[7]白雨鑫,田丽,王庆娇,等.基于机器学习的污水厂总氮超标预警及风险控制[J].中国给水排水,2025,41(19):109.
BAIYu-xin,TIANLi,WANGQing-jiao,et al.Early Warning and Risk Control for Total Nitrogen Exceeding Discharge Standard in Wastewater Treatment Plants Based on Machine Learning[J].China Water & Wastewater,2025,41(15):109.
[8]魏小岛,陈嵩钰,郭亚丽,等.基于资源一号高光谱影像的平原河网区水质遥感监测[J].中国给水排水,2026,42(7):23.
WEIXiaodao,CHENSongyu,GUOYali,et al.Remote Sensing Monitoring of Water Quality in Plain River Network Zone Based on ZY-1 Hyperspectral Imagery[J].China Water & Wastewater,2026,42(15):23.
[9]王昕妍,庞维亮,冯丽霞,等.基于可解释机器学习的污水厂碳源投加预测模型[J].中国给水排水,2026,42(7):44.
WANGXinyan,PANGWeiliang,FENGLixia,et al.Prediction Model for Carbon Source Dosing in Wastewater Treatment Plants Based on Explainable Machine Learning[J].China Water & Wastewater,2026,42(15):44.
[10]王陈波,王越,余华荣,等.基于机器学习模型优化电凝聚法去除地下水中氟化物[J].中国给水排水,2026,42(11):16.
WANGChenbo,WANGYue,YUHuarong,et al.Machine Learning-based Optimization of Electrocoagulation for Fluoride Removal from Groundwater[J].China Water & Wastewater,2026,42(15):16.