期刊文献+

基于K-means聚类与LSTM模型的多能源耦合电力负荷预测 被引量:4

Research on Multi-energy Coupled Power Load Prediction Based on K-means Clustering and LSTM Model
在线阅读 下载PDF
导出
摘要 伴随“碳达峰,碳中和”目标的提出,提升可再生能源利用率和保障能源系统灵活运用是当下电力市场发展的必然要求。与传统供能模式相比,综合能源系统考虑多能耦合协调发展,在电力市场化过程中,用能特性变化导致负荷波动规律性不明晰,影响因素的增多使负荷预测难度增大。首先分析多能耦合用能特性和影响因子间的相关性,其次对各主要因素开展K-means聚类分析,选择具有代表意义的典型日作为预测样本,采用LSTM模型预测考虑多能源间相互影响的电力负荷,建立电力负荷预测模型。最后以某综合能源园区为例进行算例分析,对比采用该方法前后预测数据的精确度,分别计算各项误差变化比例证明方法的可行性,为多能耦合的电力负荷预测提供理论基础。 With the proposal of the goal of"carbon peak,carbon neutral",improving the utilization rate of renewable energy and ensuring the flexible use of energy system are the inevitable requirements for the development requirements of the current electricity market.Compared with the traditional energy supply mode,the integrated energy system considers the coordinated development of multi-energy coupling.In the process of power marketization,the change in users'energy consumption patterns leads to an unclear regularity of load fluctuation,while the increase of influencing factors further complicates load prediction.In this paper,the correlation between the mulit-energy coupling characteristics and influence factors is analyzed,the K-means cluster analysis of the main factors is conducted,the representative typical day is selected as the prediction sample,the LSTM model is used to predict the power load,considering the interaction between multiple energy sources.Finally,a comprehensive energy park was taken as the study case.Compare with the accuracy of the prediction data before and after this method,and caculate the proportion of each error to prove the feasibility of the method,which provides a theoretical basis for the power load prediction of multi-energy coupling.
作者 葛亚明 仇晨光 谢丽荣 李艺丰 李刚 赵玉林 GE Yaming;QIU Chenguang;XIE Lirong;LI Yifeng;LI Gang;ZHAO Yulin(State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 211102,Jiangsu Province,China;NARI Group Corporation,Nanjing 211100,Jiangsu Province,China)
出处 《现代电力》 北大核心 2025年第2期369-376,共8页 Modern Electric Power
关键词 综合能源 K-MEANS聚类 LSTM模型 负荷预测 integrated energy K-means clustering LSTM model load forecasting
作者简介 葛亚明(1984),男,硕士研究生,高级工程师,研究方向为电力系统自动化,E-mail:geyam@js.sgcc.com.cn;仇晨光(1977),男,硕士研究生,高级工程师,研究方向为电力系统自动化,E-mail:cg_qiu@js.sgcc.com.cn;通信作者:谢丽荣(1980),女,硕士研究生,教授级高级工程师,研究方向为电力系统自动化,E-mail:xielirong@sgepri.sgcc.com.cn。
  • 相关文献

参考文献16

二级参考文献223

共引文献1434

同被引文献96

引证文献4

二级引证文献2

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部