基于气象因素的Stacking回归模型的短期负荷预测方法  

Shortterm Load Forecasting Method Based on Meteorological Factors Using Stacking Regression Model

作　　者：王洋 李江 张婧 格日乐图 刘秀丽 WANG Yang;LI Jiang;ZHANG Jing;Geriletu;LIU Xiuli(Inner Mongolia Power(Group)Co.,Ltd.,Hohhot 010000,China;Inner Mongolia Power(Group)Co.,Ltd.Digital Resarch Branch,Hohhot 010020,China;Inner Mongolia Power Group Mengdian Information&Telecommunication Co.,Ltd.,Hohhot 010020,China;Beijing Tsingsoft Technology Co.,Ltd.,Beijing 100085,China)

机构地区：[1]内蒙古电力(集团)有限责任公司,内蒙古呼和浩特010000 [2]内蒙古电力(集团)有限责任公司数字研究院分公司,内蒙古呼和浩特010020 [3]内蒙古电力集团蒙电信息通信产业有限责任公司,内蒙古呼和浩特010020 [4]北京清软创新科技股份有限公司,北京100085

出　　处：《电工技术》2024年第17期67-70,共4页Electric Engineering

摘　　要：为了有效提高短期负荷预测精度,针对负荷数据时序性和非线性的特点,提出了一种基于气象因素的Stacking回归模型的短期负荷预测方法。以某地区的历史日最大负荷数据、气象数据作为实验训练样本,对多种模型采用多种可行思路进行日滚动预测。经过对比分析,选择基于气象因素的Stacking回归模型作为主要预测算法,并结合相似日调整作为主要协调算法。实验结果表明,所提出的预测方法相比ARIMA模型方法、多元回归模型方法和自回归模型方法具有更高、更稳定的预测精度。Aiming at improving the accuracy of short-term load forecasting,a stacking regression model based on meteorological factors is proposed in view of the temporal and nonlinear characteristics of electricity data.Historical daily electricity data and meteorological data from a certain region are used as experimental training samples,and various feasible ideas are used for daily rolling forecasting of various models.Through comparative analysis,the Stacking regression model based on meteorological factors is selected as the main prediction algorithm,and similar day adjustment is used as the main coordination algorithm.The experimental results show that the proposed method has higher and more stable prediction accuracy than ARIMA model method,multiple regression model method,and autoregressive model method.

关 键 词：短期负荷预测 Stacking回归模型 气象因素 协调算法 

分 类 号：TM714[电气工程—电力系统及自动化]