摘要
水的供给问题是城市基础建设的重要任务之一,而对城市用水量的预测直接影响到城市的供水规划。本文对北京市2010-2020年用水量数据及其影响因素进行分析,分别建立普通多元线性回归模型和偏最小二乘回归模型对北京市用水量进行拟合和仿真预测,结果显示,偏最小二乘回归模型的系数解释更符合实际意义,且外推预测效果更好,实际应用价值较强。
Water supply is a critical undertaking in urban infrastructure development,and accurate prediction of urban water consumption is essential for effective water supply planning.In this study,we analyzed water consumption data and influencing factors in Beijing from 2010 to 2020.We developed both ordinary multivariate linear regression models and partial least squares regression models to fit and simulate the water consumption in Beijing.Our findings indicate that the coefficients derived from the partial least squares regression model provide a more meaningful interpretation,and its extrapolation prediction outperforms the ordinary multivariate linear regression model.Hence,the partial least squares regression model exhibits strong practical utility in water supply planning.
作者
曾令麒
Zeng Lingqi(School of Mathematical Sciences,South China Normal University,Guangzhou,China)
出处
《科学技术创新》
2023年第10期5-8,共4页
Scientific and Technological Innovation
关键词
用水量预测
多元线性回归
偏最小二乘回归
water consumption forecast
multiple linear regression
partial least squares regression
