电工电气杂志社

基于电动汽车充电负荷变动速率与TCN-LSTM的负荷预测

来源：电工电气发布时间：2025-06-27 13:27 浏览次数：3

基于电动汽车充电负荷变动速率与TCN-LSTM的负荷预测

汪楚皓^1，2，郭航³

(1 长沙理工大学电气与信息工程学院，湖南长沙 410114；

2 国网湖南省电力有限公司常德供电分公司，湖南常德 415000；

3 国网湖南省电力有限公司株洲供电分公司，湖南株洲 412000)

摘要：电动汽车充电负荷的随机性波动对电力系统的安全稳定性带来挑战，提出了一种基于电动汽车充电负荷变动速率与人工智能算法结合的短期预测方法。分析了电动汽车充电负荷的历史数据，提出了一种反映充电负荷速率变动特征的指标；结合时空卷积网络(TCN)和长短期记忆网络(LSTM)构建了预测模型，对充电负荷进行精准预测。实验结果表明，该方法能够有效研究区域内电动汽车用户的充电规律，对充电负荷峰谷态势的预测表现出较高的准确性，为深入分析用户充电行为模式、准确预估短期充电负荷提供了重要技术支持，对提升电力系统运行效率与稳定性具有重要意义。

关键词： 电动汽车；充电负荷；时空卷积网络；长短期记忆网络；短期负荷预测

中图分类号：TM715 ；U469.72 文献标识码：A 文章编号：1007-3175(2025)06-0019-05

Load Forecasting Based on the Variation Rate of Electric

Vehicle Charging Load and TCN-LSTM

WANG Chu-hao^1,2, GUO Hang³

（1 School of Electrical and Information Engineering, Changsha University of Science & Technology, Changsha 410114, China;

2 State Grid Hunan Electric Power Co., Ltd. Changde Power Supply Branch Company, Changde 415000, China;

3 State Grid Hunan Electric Power Co., Ltd. Zhuzhou Power Supply Branch Company, Zhuzhou 412000, China）

Abstract: The stochastic fluctuations of electric vehicle charging loads pose challenges to the safety and stability of power systems. To address this issue, this paper proposes a short-term forecasting method based on the combination of the variation rate of electric vehicle charging load and artificial intelligence algorithms. Firstly, historical data of electric vehicle charging loads are analyzed, and an indicator reflecting the variation characteristics of the charging load rate is introduced. Subsequently, a predictive model combining temporal convolutional network(TCN) and long short-term memory network(LSTM) is constructed to achieve accurate load forecasting. Experimental results demonstrate that the proposed method effectively analyzes the charging patterns of electric vehicle users in the study area and achieves high accuracy in predicting the peak-valley trends of charging loads. This study provides critical technical support for analyzing user charging behavior patterns and accurately estimating short-term charging loads, offering significant contributions to enhancing the operational efficiency and stability of power systems.

Key words: electric vehicle; charging load; temporal convolutional network; long short-term memory network; short-term load forecasting

参考文献

[1] HU Xiaosong, YUAN Hao, ZOU Changfu, et al.Coestimation of state of charge and state of health for lithium-ion batteries based on fractionalorder calculus[J].IEEE Transactions on Vehicular Technology，2018，67(11) ：10319-10329.

[2] LIU Senyi, LIU Chunhua.Generic predictive model calibration for PMSMs with different topologies[J].Green Energy and Intelligent Transportation，2022，1(1) ：100015.

[3] DANESHZAND F, COKER P J, POTTER B, et al.EV smart charging: How tariff selection influences grid stress and carbon reduction[J].Applied Energy，2023，348 ：121482.

[4] 王惠文，孟洁. 多元线性回归的预测建模方法[J]. 北京航空航天大学学报，2007，33(4) ：500-504.

[5] 彭丁聪. 卡尔曼滤波的基本原理及应用[J]. 软件导刊，2009，8(11) ：32-34.

[6] 包研科，陈然，郑宏杰，等. 指数平滑与自回归融合预测模型及实证[J] . 计算机工程与应用，2022，58(14) ：269-281.

[7] SHI Jinkai, ZHANG Weige, BAO Yan, et al.Load forecasting of electric vehicle charging stations: Attention based spatiotemporal multigraph convolutional networks[J].IEEE Transactions on Smart Grid，2024，15(3) ：3016-3027.

[8] 尤勇，孟云龙，吴景涛，等. 基于鲸鱼优化算法-支持向量回归的汽车运动状态估计[J] . 中国机械工程，2024，35(6) ：973-981.

[9] 周泽楷，侯宏娟，孙莉，等. 基于 CNN 和 BiLSTM 神经网络模型的太阳能供暖负荷预测研究[J] . 太阳能学报，2024，45(10) ：415-422.

[10] 王义国，林峰，李琦，等. 基于 TCN-LSTM 模型的电网电能质量扰动分类研究[J] . 电力系统保护与控制，2024，52(17) ：161-167.

[11] 刘宁，谢越栋，胡彬，等. CNN-LSTM 车辆运动状态识别的 AUKF 组合导航方法[J] . 中国惯性技术学报，2024，32(8) ：803-811.

[12] WANG Ning, GUO Jiahui, LIU Xiang, et al.A service demand forecasting model for oneway electric car-sharing systems combining long short-term memory networks with Granger causality test[J].Journal of Cleaner Production，2020，244 ：118812.

[13] ALI A, ZHU Y, ZAKARYA M.Exploiting dynamic spatiotemporal graph convolutional neural networks for citywide traffic flows prediction[J].Neural Networks， 2022，145 ：233-247.

[14] TIAN Jiarui, LIU Hui, GAN Wei, et al.Shortterm electric vehicle charging load forecasting based on TCN-LSTM network with comprehensive similar day identification[J].Applied Energy，2025，381 ：125174.

返回索引列表

Article retrieval

文章检索

基于电动汽车充电负荷变动速率与TCN-LSTM的负荷预测