基于嵌入式小波的CNN-LSTM电力系统负荷预测

doi:10.3969/j.issn.1008-0198.2025.01.013

湖南电力 ›› 2025, Vol. 45 ›› Issue (1): 86-92.doi: 10.3969/j.issn.1008-0198.2025.01.013

基于嵌入式小波的CNN-LSTM电力系统负荷预测

王才倩¹, 胡倩¹, 苗文捷¹, 陆沈雄²

1.杭州市电力设计院有限公司,浙江杭州 310014;
2.浙江华云信息科技有限公司,浙江杭州 310008

收稿日期:2024-10-14 修回日期:2024-11-12 出版日期:2025-02-25 发布日期:2025-03-05
通信作者: 王才倩(1987),女,高级工程师,从事电网规划、输变电工程设计、新能源接入系统设计等方面的研究工作。
基金资助:
国家电网有限公司科技项目（5100-202119574A-0-5-SF）; 浙江大有集团有限公司科技项目（DY2022-05）

CNN-LSTM Power System Load Prediction Based on Embedded Wavelet

WANG Caiqian¹, HU Qian¹, MIAO Wenjie¹, LU Shenxiong²

1. Hangzhou Electric Power Design Institute Company, Hangzhou 310014, China;
2. Zhejiang Huayun Information Technology Co., Ltd., Hangzhou 310008, China

Received:2024-10-14 Revised:2024-11-12 Online:2025-02-25 Published:2025-03-05

摘要/Abstract

摘要： 针对电力系统中负载端耗电预测精准性的问题,提出一种结合卷积神经网络和长短期记忆网络的预测方法。利用小波变换原理,对卷积神经网络中的下采样方法进行改进,使时频特征信息在下采样过程中理论上得以保留和增强。以实际耗电量数据为实验内容,结果表明,所提方法与传统方法相比,各评估指标和短周期负载预测精度方面得到显著提升。

关键词: 电力系统, 时频特征, 下采样, 嵌入式小波, 电量预测

Abstract: To address the issue of the accuracy of power consumption prediction at the load end in power systems, a prediction method combining convolutional neural networks and long short-term memory networks is proposed. By utilizing the principle of wavelet transform, the downsampling method in the convolutional neural network is improved, allowing time-frequency characteristicsinformation to be theoretically preserved and enhanced during the downsampling process. Experiments are conducted using actual power consumption data, and the results show that compared with traditional methods, the proposed method significantly improves various evaluation index and the accuracy of short-term load prediction.

Key words: power system, time-frequency characteristics, downsampling, embedded wavelets, power prediction

中图分类号:

TM715.1

王才倩, 胡倩, 苗文捷, 陆沈雄. 基于嵌入式小波的CNN-LSTM电力系统负荷预测[J]. 湖南电力, 2025, 45(1): 86-92.

WANG Caiqian, HU Qian, MIAO Wenjie, LU Shenxiong. CNN-LSTM Power System Load Prediction Based on Embedded Wavelet[J]. Hunan Electric Power, 2025, 45(1): 86-92.

参考文献

[l] 陈浩文,刘文霞,李月乔. 基于奇异谱分析与神经网络的中期负荷预测[J]. 电网技术,2020,44(4):1333-1347.
[2] 郭蕴颖,丁云峰. 基于CNN和LSTM联合预测并修正的电量缺失数据预测[J]. 计算机系统应用,2020,29(8):192-198.
[3] ZHAN X H,KOU L,XUE M T,et al.Reliable long-term energy load trend prediction model for smart grid using hierarchical decomposition self-attention network[J]. IEEE Transactions on Reliability,2023,72(2):609-621.
[4] 孙景钌,胡长洪,项烨鋆,等. 基于多核模糊C均值聚类的配电网短期负荷预测[J]. 浙江电力,2022,41(3):65-71.
[5] WANG K J,QI X,LIU H D,et al.Deep belief network based k-means cluster approach for short-term wind power forecasting[J]. Energy,2018,165:840-852.
[6] GHASVARIAN JAHROMI K,GHARAVIAN D,MAHDIANI H.A novel method for day-ahead solar power prediction based on hidden Markov model and cosine similarity[J]. Soft Computing,2020,24(7):4991-5004.
[7] 杨柳,吴延琳,张超,等. 改进最小二乘支持向量机电量预测算法[J]. 电网与清洁能源,2017,33(3):71-76.
[8] DONG Y X,MA S D,ZHANG H C,et al.Wind power prediction based on multi-class autoregressive moving average model with logistic function[J]. Journal of Modern Power Systems and Clean Energy,2022,10(5):1184-1193.
[9] 程宇也. 基于人工神经网络的短期电力负荷预测研究[D].杭州:浙江大学,2017.
[10] 陈亮,王震,王刚. 深度学习框架下LSTM网络在短期电力负荷预测中的应用[J]. 电力信息与通信技术,2017,15(5):8-11.
[11] ZHOU X,PANG C X,ZENG X H,et al.A short-term power prediction method based on temporal convolutional network in virtual power plant photovoltaic system[J]. IEEE Transactions on Instrumentation and Measurement,2023,72:9003810.
[12] LIU X Y,ZHANG Y R,ZHEN Z,et al.Spatio-temporal graph neural network and pattern prediction based ultra-short-term power forecasting of wind farm cluster[J]. IEEE Transactions on Industry Applications,2024,60(1):1794-1803.
[13] 朱凌建,荀子涵,王裕鑫,等. 基于CNN-Bi LSTM的短期电力负荷预测[J]. 电网技术,2021,45(11):4532-4539.
[14] 赵兵,王增平,纪维佳,等. 基于注意力机制的CNN-GRU短期电力负荷预测方法[J]. 电网技术,2019,43(12): 4370-4376.
[15] 任爽,杨凯,商继财,等. 基于CNN-BiGRU-Attention的短期电力负荷预测[J]. 电气工程学报,2024,19(1):344-350.
[16] YANG W Z,CHEN B,SHEN Y J,et al.WaveCNNs-AT:Wavelet-based deep CNNs of adaptive threshold for signal recognition[J]. Applied Intelligence,2023,53(23):28819-28831.
[17] LI Q F,SHEN L L,GUO S,et al.WaveCNet:wavelet integrated CNNs to suppress aliasing effect for noise-robust image classification[J]. IEEE Transactions on Image Processin,2021,30:7074-7089.
[18] LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE,1998,86(11):2278-2324.
[19] HOCHREITER S,SCHMIDHUBER J.Long short-term memory[J]. Neural Computation,1997,9:1735-1780.
[20] ZHANG R. Making convolutional networks shift-invariant again[C]//International conference on machine learning.PMLR,2019:7324-7334.

[1]	崔一阳, 黎灿兵, 潘斗南, 李振恺, 刘行. 高比例新能源电力系统中机组组合研究综述[J]. 湖南电力, 2025, 45(1): 37-45.
[2]	于志诚, 穆士才, 梁晔, 李镓辰, 林华, 陈己宸, 金鑫. 基于K-means聚类算法和BP神经网络的代理购电量预测模型研究[J]. 湖南电力, 2025, 45(1): 68-72.
[3]	赵万云, 由瑞. 基于改进蜂群算法的电力系统经济负荷分配[J]. 湖南电力, 2024, 44(5): 62-67.
[4]	于孟希, 严宇, 周展帆, 杜敏, 王炼红. 基于博弈论的负荷重分配攻击风险评估模型[J]. 湖南电力, 2024, 44(2): 1-8.
[5]	段俊峰, 李晨坤, 姚文轩, 郭思源. 基于人工智能赋能的新型电力系统关键技术综述[J]. 湖南电力, 2024, 44(1): 1-10.
[6]	张艺明, 谢荣焕, 陈孝莺. 面向新型电力系统的双有源桥变换器及其混合调制研究[J]. 湖南电力, 2023, 43(5): 16-22.
[7]	肖白, 张博. 考虑电源灵活性和互补性的多能源电力系统日前优化调度[J]. 湖南电力, 2023, 43(4): 16-23.
[8]	李帅虎, 赵翔, 蒋昀宸. 基于粒子群优化算法支持向量回归预测法的大电网电压稳定在线评估方法[J]. 湖南电力, 2022, 42(5): 22-28.

基于嵌入式小波的CNN-LSTM电力系统负荷预测

CNN-LSTM Power System Load Prediction Based on Embedded Wavelet

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 8

编辑推荐

Metrics

本文评价