摘要
短期负荷预测尤其是非节假日负荷预测对提升电力系统整体调度、支撑电网运营工作起着十分关键的作用。目前针对非节假日负荷预测的理论、方法和应用层出不穷,但是预测精度和使用范围都受到一定限制,并且经济发展对短期负荷预测的精度提出越来越高的要求,传统的机器学习算法已经难以满足人们的需求。为了提高负荷预测的精度,本文提出了利用BP神经网络进行Stacking模型融合算法,它是基于集成学习的思想,首先挑选五种预测精度较高的单模型,然后利用Stacking模型融合方法将其集成为预测精度更高的综合模型。本文采用此算法预测某省2018年非节假日负荷,结果表明该算法可以有效提高预测精度。
Short-term load forecasting,especially non-holiday load forecasting,plays a key role in improving the overall dispatch of power systems and supporting grid operations.At present,theories,methods and applications for non-holiday load forecasting are endless,but the prediction accuracy and scope of use are limited,and economic development puts higher and higher requirements on the accuracy of short-term load forecasting.Traditional machine learning algorithms are difficult to meet the needs of the people.In order to improve the accuracy of load forecasting,this paper proposes a BP neural network for Stacking model fusion algorithm.It is based on the idea of integrated learning.First,select five single models with high prediction accuracy,and then integrate them by Stacking model fusion method.A comprehensive model with a higher prediction accuracy.This paper uses this algorithm to predict the non-holiday load of a province in 2018.The results show that the algorithm is effective for improving the prediction accuracy.
作者
李昆明
厉文婕
LI Kun-ming;LI Wen-jie(Jiangsu Fangtian Electric Technology Co.,Ltd.Nanjing,Jiangsu 211102)
出处
《软件》
2019年第9期176-181,共6页
Software
