基于LSTNet-Skip的综合能源系统多元负荷超短期预测.pdfVIP

第 43 卷 第 6 期 中 国 电 机 工 程 学 报 Vol.43 No.6 Mar. 20, 2023 2023 年 3 月 20 日 Proceedings of the CSEE ©2023 Chin.Soc.for Elec.Eng. 2273 DOI ：10.13334/j.0258-8013.pcsee.213111 文章编号：0258-8013 (2023) 06-2273-10 中图分类号：TM 73 文献标识码：A 基于LSTNet-Skip 的综合能源系统 多元负荷超短期预测 * 鲁斌，霍泽健 ，俞敏 (华北电力大学计算机系，河北省 保定市 071000) Multi Load Ultra Short-term Forecasting of Integrated Energy System Based on LSTNet-Skip * LU Bin, HUO Zejian , YU Min (Department of Computer Science, North China Electric Power University, Baoding 071000, Hebei Province, China) ABSTRACT: With the development of diversified energy 比较。实验结果表明，提出的预测模型均优于其他方法且有 demand on the user side, ultra short-term load forecasting is 较高的预测精度。 very important for the planning and optimization of dynamic 关键词：综合能源系统；超短期；多元负荷预测；循环跳过； large-scale integrated energy system. Therefore, this paper 自回归 proposes a multivariate ultra short-term load forecasting model 0 引言 based on long-and short-term time-series network. First, convolutional neural network is used to extract the short-term 随着社会经济的快速发展，人们对多元能源的 dependence between multivariate loads. Then, the long-short 需求持续增加，但是由于化石能源逐渐枯竭，传统 term memory network is used to capture the long term 的火力发电效率低下、环境污染严重等问题，综合 dependence of load sequence, and the ultra long-term repetitive pattern of