西北工程技术学报2024，Vol.23Issue(4)：330-335,346,7.
基于GM-LSSVR的宁夏工业用水预测
Prediction of Industrial Water Consumption in Ningxia Based on GM-LSSVR
李淑霞 1王生鑫 1张海涛 2童钰芳 1马静 1赵宇翔2
作者信息
- 1. 宁夏回族自治区水文水资源监测预警中心,宁夏 银川 750000
- 2. 中国水利水电科学研究院,北京 100044
- 折叠
摘要
Abstract
Accurate prediction of future industrial water consumption is of great significance for Ningxia's industrial development planning and sustainable economic and social development.The GM-LSSVR combination model based on HP filter decomposition can overcome the limitations of using a single prediction model and improve prediction accuracy.Based on data from 2004 to 2021 on industrial water consumption,economic indicators,comprehensive physical quantity index,industrial physical quantity index and other related data in Ningxia,the main influencing factors of industrial water consumption are identified using grey correlation analysis and Pearson correlation analysis.On this basis,the combined model is employed to predict Ningxia's future industrial water consumption,and the accuracy and rationality of the forecast results are analyzed.The results indicate that the proportion of manufacturing value added to total industrial value added,total energy consumption,thermal power generation,diesel consumption,natural gas consumption,and electricity consumption are the main factors affecting the industrial water consumption in Ningxia.The GM-LSSVR combination model based on HP filter decomposition demonstrates good applicability and prediction performance in forecasting industrial water consumption in Ningxia.The model predicts that industrial water consumption in Ningxia will reach 453 million m3 in 2025.关键词
工业用水/用水预测/灰色模型/最小二乘支持向量回归模型/HP滤波分解Key words
industrial water consumption/water use forecasting/grey model/LSSVR/HP filter decomposition分类
建筑与水利引用本文复制引用
李淑霞,王生鑫,张海涛,童钰芳,马静,赵宇翔..基于GM-LSSVR的宁夏工业用水预测[J].西北工程技术学报,2024,23(4):330-335,346,7.
