基于互信息和LSTM的用户负荷短期预测

doi:10.12204/j.issn.1000-7229.2022.07.011

电力建设 ›› 2022, Vol. 43 ›› Issue (7): 96-102.doi: 10.12204/j.issn.1000-7229.2022.07.011

基于互信息和LSTM的用户负荷短期预测

钟劲松¹, 王少林², 冉懿¹(), 冉新涛², 于金平², 俞海猛³

1.国网新疆电力有限公司电力科学研究院,乌鲁木齐市 830000
2.国网奎屯供电公司,新疆维吾尔自治区奎屯市 833200
3.国电南瑞南京控制系统有限公司,南京市 211106

收稿日期:2021-10-14 出版日期:2022-07-01 发布日期:2022-06-30
通讯作者: 冉懿 E-mail:759702357@qq.com.cn
作者简介:钟劲松(1969),女,高级工程师,研究方向为电力营销管理及电能计量;
王少林(1972),男,工程师,主要从事电力营销管理及电能计量研究工作;
冉新涛(1983),男,工程师,主要从事电力计量及电力仪器仪表研究工作;
于金平(1987),男,工程师,研究方向为电力营销、电能计量、线损管理等;
俞海猛(1988),男,工程师,研究方向为用电信息采集和电能计量。
基金资助:
国家重点研发计划资助“城区用户与电网供需友好互动系统”(2016YFB0901100)

Short-Term Consumer Load Forecasting Based on Mutual Information and LSTM

ZHONG Jingsong¹, WANG Shaolin², RAN Yi¹(), RAN Xintao², YU Jinping², YU Haimeng³

1. State Grid Xinjiang Electric Power Co., Ltd. Electric Power Research Institute, Urumqi 830000, China
2. State Grid Kuitun Power Supply Company,Kuitun 833200,Xinjiang Uygur Autonomous Region,China
3. NARI-TECH Nanjing Control System Co., Ltd.,Nanjing 211106,China

Received:2021-10-14 Online:2022-07-01 Published:2022-06-30
Contact: RAN Yi E-mail:759702357@qq.com.cn
Supported by:
National Key Research and Development Program of China(2016YFB0901100)

摘要/Abstract

摘要：

相对于系统级负荷,用户负荷具有基数小、波动性与随机性更强的特点,加大了用户负荷预测的难度。文章借助互信息与深度学习理论,提出了一种基于最大相关最小冗余(max-relevance and min-redundancy, mRMR)和长短期记忆网络(long-short term memory networks, LSTM)的用户负荷短期预测模型。首先,采用mRMR算法对特征变量进行排序并选取合适的输入变量集合,mRMR既可以保证输入变量与目标值间互信息值最大,又使得变量间冗余性最小。接着,对选取的输入变量集合建立LSTM预测模型,LSTM能较好处理和预测延迟较长的时间序列,且不会存在梯度消失和梯度爆炸现象。最后,通过算例验证了所提算法的有效性。

关键词: 用户负荷短期预测, 互信息, 最大相关最小冗余算法(mRMR), 长短期记忆网络(LSTM)

Abstract:

Compared with the system-level load, consumer load has the characteristics of small base, stronger volatility and randomness, which increases the difficulty of consumer load forecasting. With the help of mutual information and deep learning theory, this paper proposes a short-term consumer load forecasting model based on max-relevance and min-redundancy (mRMR) and long short-term memory network (LSTM). Firstly, the mRMR algorithm is used to sort the characteristic variables and select a suitable set of input variables. mRMR can not only ensure the maximum mutual information value between the input variable and the target value, but also minimize the redundancy between the variables. Secondly, the LSTM forecasting model is established for the selected set of input variables. LSTM can better process and forecast time series with long delays, and there will be no gradient disappearance and gradient explosion. Finally, an example is used to verify the effectiveness of the algorithm in this paper.

Key words: short-term consumer load forecasting, mutual information, max-relevance and min-redundancy algorithm, long short-term memory network

中图分类号:

TM715

钟劲松, 王少林, 冉懿, 冉新涛, 于金平, 俞海猛. 基于互信息和LSTM的用户负荷短期预测[J]. 电力建设, 2022, 43(7): 96-102.

ZHONG Jingsong, WANG Shaolin, RAN Yi, RAN Xintao, YU Jinping, YU Haimeng. Short-Term Consumer Load Forecasting Based on Mutual Information and LSTM[J]. ELECTRIC POWER CONSTRUCTION, 2022, 43(7): 96-102.

图/表 9

图1 LSTM 结构模型

Fig. 1 LSTM structure model

图2 用户负荷不同模型预测误差变化曲线

Fig.2 Forecasting error curves of different models of consumer load

图3 用户负荷真实值及模型预测值(2016年8月9日)

Fig.3 True value of user load and model prediction value (August 9th, 2016)

表1 用户负荷提前24 h预测误差统计(2016年8月9日)

Table 1 Error statistics of consumer load forecasting 24 hours in advance (August 9th,2016)

图4 用户负荷真实值及提前24 h预测值(2016年11月20—26日)

Fig.4 Real value of user load and the predicted value 24 h in advance(Nov.20—26, 2016)

图5 用户负荷真实值及提前1 h预测值(2016年11月20—26日)

Fig.5 The real value of user load and the predicted value 1h in advance(Nov.20—26, 2016)

表2 用户负荷提前24 h及提前1 h预测误差统计(2016年11月20—26日)

Table 2 Error statistics of consumer load forecasting 24 hours and 1 hour in advance (Nov.20—26, 2016)

图6 提前24 h用户负荷预测误差分布

Fig.6 Error distribution of consumer load forecasting 24 hours in advance

图7 提前1 h用户负荷预测误差分布

Fig.7 The error distribution of consumer load forecasting 1 hour in advance

参考文献 21

[1]	张智, 卢峰, 林振智, 等. 考虑用户有限理性的售电公司峰谷组合电力套餐设计[J]. 电力系统自动化, 2021, 45(16): 114-123.
	ZHANG Zhi, LU Feng, LIN Zhenzhi, et al. Peak-valley combination electricity package design for electricity retailer considering bounded rationality of consumers[J]. Automation of Electric Power Systems, 2021, 45(16): 114-123.
[2]	高瑞, 郭红霞, 杨苹, 等. 基于用户侧可调节资源的售电公司实时市场偏差电量管控策略[J]. 电力建设, 2019, 40(6): 114-122.
	GAO Rui, GUO Hongxia, YANG Ping, et al. Research on control strategy of spot market deviation of electricity sales company based on user-side adjustable resources[J]. Electric Power Construction, 2019, 40(6): 114-122.
[3]	徐杰彦, 许雯旸, 褚渊, 等. 区域尺度住宅建筑日用电负荷模型构建方法研究[J]. 中国电力, 2020, 53(8): 29-39.
	XU Jieyan, XU Wenyang, CHU Yuan, et al. Residential electricity load model construction in district scale[J]. Electric Power, 2020, 53(8): 29-39.
[4]	叶圣永, 魏俊, 阮贺彬, 等. 多样化柔性负荷深度互动下的虚拟电厂分布鲁棒购售电策略[J]. 电力建设, 2021, 42(1): 59-66.
	YE Shengyong, WEI Jun, RUAN Hebin, et al. Distributional robust electricity trading strategy of virtual power plant with deep interaction of diversified flexible loads[J]. Electric Power Construction, 2021, 42(1): 59-66.
[5]	白登辉. 计及偏差电量考核机制的人工神经网络售电量预测模型[J]. 电工电能新技术, 2020, 39(6): 58-66.
	BAI Denghui. Artificial neural network electricity sales forecast model considering deviation electricity assessment mechanism[J]. Advanced Technology of Electrical Engineering and Energy, 2020, 39(6): 58-66.
[6]	顾洁, 孟璐, 郑睿程, 等. 考虑集群辨识的海量用户负荷分层概率预测[J]. 电力系统自动化, 2021, 45(5): 71-78.
	GU Jie, MENG Lu, ZHENG Ruicheng, et al. Load-stratified probability forecasting for massive users considering cluster identification[J]. Automation of Electric Power Systems, 2021, 45(5): 71-78.
[7]	杨德州, 魏勇, 李万伟, 等. 基于多层分解-累加原理的城市综合体月度用电量预测方法[J]. 电力建设, 2021, 42(2): 27-34.
	YANG Dezhou, WEI Yong, LI Wanwei, et al. Multi-layer decomposition-accumulation principle-based monthly electricity consumption forecasting method for urban complex[J]. Electric Power Construction, 2021, 42(2): 27-34.
[8]	孙庆凯, 王小君, 张义志, 等. 基于LSTM和多任务学习的综合能源系统多元负荷预测[J]. 电力系统自动化, 2021, 45(5): 63-70.
	SUN Qingkai, WANG Xiaojun, ZHANG Yizhi, et al. Multiple load prediction of integrated energy system based on long short-term memory and multi-task learning[J]. Automation of Electric Power Systems, 2021, 45(5): 63-70.
[9]	IMANI M. Electrical load-temperature CNN for residential load forecasting[J]. Energy, 2021, 227: 120480. doi: 10.1016/j.energy.2021.120480 URL
[10]	FUJIMOTO Y, FUJITA M, HAYASHI Y. Deep reservoir architecture for short-term residential load forecasting: An online learning scheme for edge computing[J]. Applied Energy, 2021, 298: 117176. doi: 10.1016/j.apenergy.2021.117176 URL
[11]	SUN M Y, ZHANG T Q, WANG Y, et al. Using Bayesian deep learning to capture uncertainty for residential net load forecasting[J]. IEEE Transactions on Power Systems, 2020, 35(1): 188-201. doi: 10.1109/TPWRS.2019.2924294 URL
[12]	周冰钰, 刘博, 王丹, 等. 基于自组织中心K-means算法的用户互动用电行为聚类分析[J]. 电力建设, 2019, 40(1): 68-76.
	ZHOU Bingyu, LIU Bo, WANG Dan, et al. Clustering analysis of user power interaction behavior based on self-organizing center K-means algorithm[J]. Electric Power Construction, 2019, 40(1): 68-76.
[13]	LIANG Y, NIU D X, HONG W C. Short term load forecasting based on feature extraction and improved general regression neural network model[J]. Energy, 2019, 166: 653-663. doi: 10.1016/j.energy.2018.10.119 URL
[14]	LIU Y, WANG W, GHADIMI N. Electricity load forecasting by an improved forecast engine for building level consumers[J]. Energy, 2017, 139: 18-30. doi: 10.1016/j.energy.2017.07.150 URL
[15]	HUANG N T, HU Z Q, CAI G W, et al. Short term electrical load forecasting using mutual information based feature selection with generalized minimum-redundancy and maximum-relevance criteria[J]. Entropy, 2016, 18(9): 330. doi: 10.3390/e18090330 URL
[16]	朱乔木, 李弘毅, 王子琪, 等. 基于长短期记忆网络的风电场发电功率超短期预测[J]. 电网技术, 2017, 41(12): 3797-3802.
	ZHU Qiaomu, LI Hongyi, WANG Ziqi, et al. Short-term wind power forecasting based on LSTM[J]. Power System Technology, 2017, 41(12): 3797-3802.
[17]	KONG W C, DONG Z Y, JIA Y W, et al. Short-term residential load forecasting based on LSTM recurrent neural network[J]. IEEE Transactions on Smart Grid, 2019, 10(1): 841-851. doi: 10.1109/TSG.2017.2753802 URL
[18]	ZANG H X, XU R Q, CHENG L L, et al. Residential load forecasting based on LSTM fusing self-attention mechanism with pooling[J]. Energy, 2021, 229: 120682. doi: 10.1016/j.energy.2021.120682 URL
[19]	明彤彤, 赵晶, 王晓磊, 等. 基于改进LSTM的脉冲大倍率工况下锂电池SOC估计[J]. 电力系统保护与控制, 2021, 49(8): 144-150.
	MING Tongtong, ZHAO Jing, WANG Xiaolei, et al. SOC estimation of a lithium battery under high pulse rate condition based on improved LSTM[J]. Power System Protection and Control, 2021, 49(8): 144-150.
[20]	葛磊蛟, 羡一鸣, 王中冠, 等基于GWO-GRNN的区域配电网短期负荷预测模型[J]. 中国电机工程学报, 2021, 7(5): 1093-1101.
	GE Leijiao, XIAN Yiming, WANG Zhongguan, et al. Short-term load forecasting of regional distribution network based on generalized regression neural network optimized by grey wolf optimization algorithm[J]. CSEE Journal of Power and Energy Systems, 2021, 7(5): 1093-1101.
[21]	李相俊, 许格健. 基于长短期记忆神经网络的风力发电功率预测方法[J]. 发电技术, 2019, 40(5): 426-433.
	LI Xiangjun, XU Gejian. Wind power prediction method based on long short-term memory neural network[J]. Power Generation Technology, 2019, 40(5): 426-433.

基于互信息和LSTM的用户负荷短期预测

Short-Term Consumer Load Forecasting Based on Mutual Information and LSTM

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 21

相关文章 5

编辑推荐

Metrics

本文评价

[1]	王哲, 万宝, 凌天晗, 董晓红, 穆云飞, 邓友均, 唐舒懿. 基于谱聚类和LSTM神经网络的电动公交车充电负荷预测方法[J]. 电力建设, 2021, 42(6): 58-66.
[2]	朱庆, 郑红娟, 唐子逸, 韦思雅, 邹子骁, 吴熙. 基于生成对抗网络的综合能源负荷场景生成方法[J]. 电力建设, 2021, 42(12): 1-8.
[3]	陈倩，齐林海，王红. 基于LSTM网络的谐波多标签分类[J]. 电力建设, 2019, 40(7): 101-106.
[4]	杨甲甲, 刘国龙, 赵俊华, 文福拴, 董朝阳. 采用长短期记忆深度学习模型的工业负荷短期预测方法[J]. 电力建设, 2018, 39(10): 20-27.
[5]	李刚，杨立业，刘福炎，俞敏，宋雨，文福拴. 能源互联网关联数据融合的互信息方法[J]. 电力建设, 2016, 37(9): 22-.