自动化与仪器仪表2024,Issue(4) :149-153.DOI:10.14016/j.cnki.1001-9227.2024.04.149

综合能源接入下多元电力负荷短期预测方法

Short-term prediction method of multiple power load under integrated energy access

胡志强 祝君剑 常凯旋 肖园
自动化与仪器仪表2024,Issue(4) :149-153.DOI:10.14016/j.cnki.1001-9227.2024.04.149

综合能源接入下多元电力负荷短期预测方法

Short-term prediction method of multiple power load under integrated energy access

胡志强 1祝君剑 2常凯旋 2肖园3
扫码查看

作者信息

  • 1. 国网江西省电力有限公司供用电部,南昌 330013
  • 2. 国网江西省电力有限公司供电服务管理中心,南昌 330001
  • 3. 国网江西省电力有限公司经济技术研究院,南昌 330012
  • 折叠

摘要

多元电力负荷之间的变化情况影响着综合能源系统的正常运行和协调管理.因此,提出综合能源接入下多元电力负荷短期预测方法.采用基于Copula理论,分析多元电力负荷之间和影响因素之间的非线性特性,形成多变量时间序列.将其输入最小二乘支持向量机模型中,完成多元电力负荷短期预测.采用果蝇优化算法优化预测模型,获取最佳预测结果,实现多元电力负荷短期预测.实验结果表明,当惩罚系数为0.4,搜索距离为5时,可实现最佳多元电力负荷短期预测,预测结果的均方误差为0.6%~2.5%,说明所提方法的预测效果较好,为综合能源系统的运行管理提供了支撑.

Abstract

The changes between multiple power loads affect the normal operation and coordinated management of the integrated energy system.Therefore,a short-term prediction method for multiple power loads under comprehensive energy access is proposed.Using Copula theory,analyze the nonlinear characteristics between multiple power loads and influencing factors,and form a multivari-ate time series.Input it into the least squares support vector machine model to complete short-term prediction of multivariate power loads.Using the fruit fly optimization algorithm to optimize the prediction model,obtain the best prediction results,and achieve short-term prediction of multiple power loads.The experimental results show that when the penalty coefficient is 0.4 and the search dis-tance is 5,the optimal short-term prediction of multiple power loads can be achieved,with a mean square error of 0.6%-2.5%.This indicates that the proposed method has good prediction performance and provides support for the operation and management of the comprehensive energy system.

关键词

综合能源接入/多元电力负荷/短期预测/耦合特性/时间序列/非线性

Key words

integrated energy access/multiple power loads/short term forecast/coupling characteristics/time series/nonlinear

引用本文复制引用

出版年

2024
自动化与仪器仪表
重庆工业自动化仪表研究所,重庆市自动化与仪器仪表学会

自动化与仪器仪表

CSTPCD
影响因子:0.327
ISSN:1001-9227
参考文献量19
段落导航相关论文