首页|基于指数平滑的ARIMA模型预测河流水质研究

基于指数平滑的ARIMA模型预测河流水质研究

扫码查看
依托于上海市嘉定区横沥河2009年~2022年高锰酸盐指数和总磷数据,构建ARIMA模型并预测未来数值,加入指数平滑预处理,对比处理前后ARIMA模型表现.结果表明:ARIMA模型对高锰酸盐指数预测结果相对较为准确,对总磷预测结果相对误差较大.指数平滑预处理能增强ARIMA模型拟合度和预测精度,相较于单一ARIMA模型,指数平滑-ARIMA模型表现明显改善,可将高锰酸盐指数和总磷预测模型拟合优度R2分别提高0.511和0.397,显著降低总磷预测值相对误差,最大可减少52.94%.
Prediction of River Water Quality by ARIMA Model Based on Exponential Smoothing
Based on the permanganate index and total phosphorus data of Hengli River,Jiading District,Shanghai from 2009 to 2022,an ARIMA model was constructed and the future values were predicted.Exponential smoothing pretreatment was added to compare the performance of the ARIMA model before and after the treatment.The results show that the ARIMA model is relatively accurate in predicting permanganate index and has a large error in predicting total phosphorus.Compared with the single ARIMA model,the performance of the exponential smooth-ARIMA model was significantly improved,and the permanganate index and the fit excellence R2 of the total phosphorus prediction model were increased by 0.511 and 0.397,respectively,and the relative error of the total phosphorus prediction value was significantly reduced.The maximum reduction is 52.94%.

environmentriver water quality predictionARIMAexponential smoothingpermanganate indextotal phosphorus

张树艳、陈奇兵、蔡怡洁

展开 >

上海市嘉定区环境监测站,上海 201800

环境 河流水质预测 ARIMA 指数平滑 高锰酸盐指数 总磷

2024

广东化工
广东省石油化工研究院

广东化工

影响因子:0.288
ISSN:1007-1865
年,卷(期):2024.51(6)
  • 12