摘要
为了充分挖掘电力负荷与多维特征因素的非线性关系,提高负荷预测精度,提出了一种基于随机森林和双向长短期记忆(bidirectional long-short-term memory,Bi-LSTM)网络的超短期负荷预测方法。首先,采用基于随机森林的特征选择算法,筛选与负荷关联性强的特征作为预测模型的输入;其次,构建Bi-LSTM网络,对特征选择后的负荷数据进行预测;最后,采用某市真实负荷数据进行仿真分析。结果表明,所提方法与传统预测方法相比,具有更高的预测精度,为精确预测具有多维特征因素的负荷提供参考。
In order to fully explore the non-linear relationship between power load and multi-dimensional feature factors and improve the accuracy of load forecasting,an ultra-short-term load forecasting method based on random forest and bidirectional long short-term memory(Bi-LSTM)network was proposed.Firstly,use the feature selection algorithm based on random forest to screen the features with strong correlation with the load as the input of the prediction model;secondly,construct the Bi-LSTM network to predict the load data after feature selection;finally,use real load data of a certain city for simulation analysis.The results show that the method proposed in this article has higher prediction accuracy than traditional prediction methods,and provides a reference for accurately predicting the load with multi-dimensional feature factors.
作者
伍乙杰
黄文灏
赖仕达
陈光宇
贾鹏
李家兴
Wu Yijie;Huang Wenhao;Lai Shida;Chen Guangyu;Jia Peng;Li Jiaxing(School of Electric Power Engineering,Nanjing Institute of Technology,Nanjing Jiangsu 211167,China;State Grid Fujian Electric Power Co.,Ltd.,Sanming Power Supply Branch,Sanming Fujian 353000,China)
出处
《电气自动化》
2022年第5期35-37,40,共4页
Electrical Automation
基金
南京工程学院2021年大学生科技创新基金项目(TB202117043)。
关键词
多维特征
负荷预测
随机森林
双向长短期记忆网络
特征选择
multidimensional characteristics
load forecasting
random forest
bidirectional long-short-term memory(Bi-LSTM)
feature selection
作者简介
伍乙杰(1996-),男,江苏宿迁人,硕士研究生,研究方向为电力系统运行与控制。
