考虑动态时间锚点和典型特征约束的年日均负荷曲线预测

doi:10.11930/j.issn.1004-9649.202308114

中国电力 ›› 2024, Vol. 57 ›› Issue (11): 36-47.DOI: 10.11930/j.issn.1004-9649.202308114

考虑动态时间锚点和典型特征约束的年日均负荷曲线预测

李丹¹(), 贺帅¹(), 颜伟², 胡越³, 方泽仁³, 梁云嫣³

1. 三峡大学电气与新能源学院，湖北宜昌　443002
2. 重庆大学电气工程学院，重庆　400044
3. 新能源微电网湖北省协同创新中心，湖北宜昌　443002

收稿日期:2023-08-28 接受日期:2024-02-05 出版日期:2024-11-28 发布日期:2024-11-27
作者简介:李丹（1980—），女，通信作者，博士，副教授，从事负荷和新能源功率预测、电力系统不确定性分析、电力系统优化运行研究，E-mail：lucy2140@163.com
贺帅（1997—），男，硕士研究生，从事电力系统运行与控制研究，E-mail：christmashs@163.com
基金资助:
国家自然科学基金资助项目（考虑气象因素不确定性与数据降维技术的中期小时级源荷功率场景概率预测方法研究，51807109）。

Annual Daily Average Load Curve Prediction Considering Dynamic Time Anchors and Typical Feature Constraints

Dan LI¹(), Shuai HE¹(), Wei YAN², Yue HU³, Zeren FANG³, Yunyan LIANG³

1. College of Electric Engineering and New Energy, China Three Gorges University, Yichang 443002, China
2. School of Electrical Engineering, Chongqing University, Chongqing 400044, China
3. Hubei Provincial Collaborative Innovation Center for New Energy Microgrid, Yichang 443002, China

Received:2023-08-28 Accepted:2024-02-05 Online:2024-11-28 Published:2024-11-27
Supported by:
This work is supported by National Natural Science Foundation of China (Research on Mid-term Source & Load Hourly Power Scenario Probabilistic Prediction Method Considering Uncertainty of Climatic Factors and Data Dimensionality Reduction Technology, No.51807109).

摘要/Abstract

摘要：

基于负荷趋势性、周期性和受日历特征影响的特点，考虑动态时间锚点和典型特征约束，实现年日均负荷曲线精确预测。首先，根据历史和预测年的日历关联关系建立动态时间锚点矩阵，结合标幺化和周期平滑处理后的历史年日均负荷形状因子曲线，提出DTA-Soft-DBA方法以获得预测年的日均负荷形状因子预测曲线；然后，进行反标幺化和反周期平滑处理，并结合电力电量特征预测值进行典型特征约束修正，获得年日均负荷预测曲线。基于某地区的算例结果表明，所提方法具有更高的预测精度，其结果与典型特征预测值相吻合，符合年内时序变化规律，能有效整合具有不同日历特征的历史样本时序共性规律。

关键词: 负荷预测, 年日均负荷曲线, Soft-DBA, 特征约束, 日历特征

Abstract:

Based on the trends, periodicity and calendar features of power load, it is realized to accurately predict the annual daily average load curves considering the dynamic time anchors and typical feature constraints, Firstly, a dynamic time anchor matrix is built based on the calendar association between the historical year and the target year. Then, based on the historical annual daily average load shape-factor curves obtained after normalization and periodic smoothing treatment, it is proposed to use the DTA-Soft-DBA to predict the target year's daily average load shape-factor curve. After inverse normalization and inverse periodic smoothing treatment, the annual daily average load prediction curves are obtained by correcting the typical feature constraints with the predicted value of power and electricity features. The case study results of an area in China show that the proposed method has higher prediction accuracy, and the results are consistent with the predicted values of typical features and the temporal variations within the target year. The proposed method can effectively integrate the common rules of historical sample time series with different calendar characteristics, which is reasonable and interpretable.

Key words: load forecasting, annual daily average load curve, Soft-DBA, feature constraints, calendar features

李丹, 贺帅, 颜伟, 胡越, 方泽仁, 梁云嫣. 考虑动态时间锚点和典型特征约束的年日均负荷曲线预测[J]. 中国电力, 2024, 57(11): 36-47.

Dan LI, Shuai HE, Wei YAN, Yue HU, Zeren FANG, Yunyan LIANG. Annual Daily Average Load Curve Prediction Considering Dynamic Time Anchors and Typical Feature Constraints[J]. Electric Power, 2024, 57(11): 36-47.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202308114

https://www.electricpower.com.cn/CN/Y2024/V57/I11/36

图/表 12

图 1 基于FC-DTA-Soft-DBA的年日均负荷曲线预测总体框架

Fig.1 The overall framework for annual daily average load curve forecasting based on FC-DTA-Soft-DBA

图 2 2015年对2020年春节影响权重惩罚矩阵

Fig.2 Weighted impact penalty matrix of 2015 to 2020 for the spring festival

图 3 2015年对2020年时间锚点矩阵的子影响区域

Fig.3 Sub-influence areas of the time anchor matrix of 2015 to 2020

图 4 DTA-Soft-DBA的负荷形状因子曲线预测流程

Fig.4 Load shape-factor curve prediction flow chart based on Soft-DBA with dynamic time anchor points

图 5 2015—2021年的日均负荷及年均负荷曲线

Fig.5 Daily and annual average load curves from 2015 to 2021

图 6 2015—2019年的日均负荷曲线

Fig.6 Average load curves from 2015 to 2019

图 7 2015—2019年的年日均负荷和节假日日均负荷曲线

Fig.7 The annual daily average load and holiday daily average load curves from 2015 to 2019

图 8 不同方法的日均负荷形状因子预测曲线对比

Fig.8 Comparison of daily average load shape-factor prediction curves by different methods

图 9 不同方法的2021年日均负荷预测曲线对比

Fig.9 Comparison of daily average load prediction curves by different methods for 2021

表 1 不同方法的2020年和2021年评价指标比较

Table 1 Evaluating indicator comparison of different methods for 2020 and 2021

方法	2020			2021			平均指标
方法	E_MAPE/%	E_RMSE/MW	R²	E_MAPE/%	E_RMSE/MW	R²	E_MAPE/%	E_RMSE/MW	R²
Holt-Winters	14.47	1856.94	0.28	8.51	1555.69	0.44	11.49	1706.32	0.36
GM	12.45	1803.01	0.32	11.28	2012.11	0.06	11.86	1907.56	0.19
Naive	12.00	1773.68	0.34	11.74	2172.00	–0.09	11.87	1972.84	0.13
ED	11.98	1685.25	0.4	8.32	1662.38	0.36	10.15	1673.82	0.38
DBA	12.32	1766.68	0.35	8.56	1714.12	0.32	10.44	1740.40	0.34
SGD-DBA	12.17	1728.36	0.37	8.51	1694.65	0.34	10.34	1711.51	0.36
Soft-DDBA	11.10	1477.22	0.54	7.90	1480.13	0.49	9.50	1478.68	0.52
FC-DTA-Soft-DBA	6.87	1054.16	0.77	6.79	1591.24	0.41	6.83	1322.70	0.59

图 10 2020、2021年FC-DTA-Soft-DBA消融实验的预测误差均值对比

Fig.10 Mean error comparison of FC-DTA-Soft-DBA ablation experiments in 2020 and 2021

图 11 2021年消融实验预测结果的最值和月均值负荷对比

Fig.11 Comparison of annual max/min and monthly average prediction load of ablation experiments in 2021

参考文献 25

1	LÓPEZ J C, RIDER M J, WU Q W. Parsimonious short-term load forecasting for optimal operation planning of electrical distribution systems[J]. IEEE Transactions on Power Systems, 2019, 34 (2): 1427- 1437. DOI
2	张籍, 薛儒涛, 刘慧, 等. 基于深度信念网络的不同行业中长期负荷预测[J]. 电力系统及其自动化学报, 2019, 31 (9): 12- 19, 27.
	ZHANG Ji, XUE Rutao, LIU Hui, et al. Medium-and long-term load forecasting for different industries based on deep belief network[J]. Proceedings of the CSU-EPSA, 2019, 31 (9): 12- 19, 27.
3	LUSIS P, KHALILPOUR K R, ANDREW L, et al. Short-term residential load forecasting: impact of calendar effects and forecast granularity[J]. Applied Energy, 2017, 205, 654- 669. DOI
4	SINGH P, DWIVEDI P. A novel hybrid model based on neural network and multi-objective optimization for effective load forecast[J]. Energy, 2019, 182, 606- 622. DOI
5	SADEGHI M S. A hybrid intelligent-based linear-nonlinear model for accurate daily peak load forecasting[J]. Journal of Intelligent & Fuzzy Systems, 2014, 26 (6): 3013- 3020.
6	祝青, 孟祥娟, 陈晓峰, 等. 基于时间尺度组合优化的中长期负荷预测方法[J]. 大众用电, 2021, 36 (6): 20- 22.
7	ZHU K D, LI Y P, MAO W B, et al. LSTM enhanced by dual-attention-based encoder-decoder for daily peak load forecasting[J]. Electric Power Systems Research, 2022, 208, 107860. DOI
8	ZHENG H T, YUAN J B, CHEN L. Short-term load forecasting using EMD-LSTM neural networks with a xgboost algorithm for feature importance evaluation[J]. Energies, 2017, 10 (8): 1168. DOI
9	SUN T H, ZHANG T Y, TENG Y, et al. Monthly electricity consumption forecasting method based on X12 and STL decomposition model in an integrated energy system[J]. Mathematical Problems in Engineering, 2019, 2019, 9012543.
10	TRULL O, GARCÍA-DÍAZ J C, PEIRÓ-SIGNES A. Multiple seasonal STL decomposition with discrete-interval moving seasonalities[J]. Applied Mathematics and Computation, 2022, 433, 127398. DOI
11	全一鸣, 喻敏, 王文波, 等. 基于分形优化的VMD和GA-BP的短期风速预测[J]. 太阳能学报, 2023, 44 (7): 436- 446.
	QUAN Yiming, YU Min, WANG Wenbo, et al. Short-term wind speed prediction based on fractal optimization of VMD-GA-BP[J]. Acta Energiae Solaris Sinica, 2023, 44 (7): 436- 446.
12	CHEN Y T, ZHANG D X. Theory-guided deep-learning for electrical load forecasting (TgDLF) via ensemble long short-term memory[J]. Advances in Applied Energy, 2021, 1, 100004. DOI
13	张颖超, 沈子豪, 马伟叁, 等. 相似样本优化选取的短期风电功率预测[J]. 重庆理工大学学报(自然科学), 2022, (1): 277- 284.
	ZHANG Yingchao, SHEN Zihao, MA Weisan, et al. Short-term wind power forecast based on DTW similar optimization sample selection[J]. Journal of Chongqing University of Technology (Natural Science), 2022, (1): 277- 284.
14	LIANG H S, MA J. Develop load shape dictionary through efficient clustering based on elastic dissimilarity measure[J]. IEEE Transactions on Smart Grid, 2021, 12 (1): 442- 452. DOI
15	CUTURI M, BLONDEL M. Soft-DTW: a differentiable loss function for time-series[C]//Proceedings of the 34th International Conference on Machine Learning-Volume 70. Sydney, NSW, Australia. ACM, 2017: 894–903.
16	PETITJEAN F, KETTERLIN A, GANÇARSKI P. A global averaging method for dynamic time warping, with applications to clustering[J]. Pattern Recognition, 2011, 44 (3): 678- 693. DOI
17	WANG B, TAI N L, ZHAI H Q, et al. A new ARMAX model based on evolutionary algorithm and particle swarm optimization for short-term load forecasting[J]. Electric Power Systems Research, 2008, 78 (10): 1679- 1685. DOI
18	丁义, 杨建. 欧式距离与标准化欧式距离在k近邻算法中的比较[J]. 软件, 2020, 41 (10): 135- 136, 140. DOI
	DING Yi, YANG Jian. Comparison between euclidean distance and standardized euclidean distance in k-nearest neighbor algorithm[J]. Computer Engineering & Software, 2020, 41 (10): 135- 136, 140. DOI
19	苏振宇, 龙勇, 汪於. 基于季节调整和Holt-Winters的月度负荷预测方法[J]. 中国管理科学, 2019, 27 (3): 30- 40.
	SU Zhenyu, LONG Yong, WANG Yu. A hybrid monthly load forecasting method based on seasonal adjustment and holt-winters[J]. Chinese Journal of Management Science, 2019, 27 (3): 30- 40.
20	任峰, 李伟, 丁超. 基于灰色-周期外延组合模型的电力负荷预测[J]. 电网技术, 2007, 31 (24): 52- 54.
	REN Feng, LI Wei, DING Chao. Load forecasting based on gray-periodic extensional combinatorial model[J]. Power System Technology, 2007, 31 (24): 52- 54.
21	SCHULTZ D, JAIN B. Nonsmooth analysis and subgradient methods for averaging in dynamic time warping spaces[J]. Pattern Recognition, 2018, 74, 340- 358. DOI
22	ZHANG S Y, CHEN R H, CAO J C, et al. A CNN and LSTM-based multi-task learning architecture for short and medium-term electricity load forecasting[J]. Electric Power Systems Research, 2023, 222, 109507. DOI
23	RENTACHINTALA L R N P, MUTUKURU GANGIREDDY M R, MOHAPATRA P K. Trends of surface water quality of the Krishna River, India during the urbanization process[J]. Environmental Quality Management, 2023, 32 (3): 111- 121. DOI
24	张倩, 马愿, 李国丽, 等. 频域分解和深度学习算法在短期负荷及光伏功率预测中的应用[J]. 中国电机工程学报, 2019, 39 (8): 2221- 2230.
	ZHANG Qian, MA Yuan, LI Guoli, et al. Applications of frequency domain decomposition and deep learning algorithms in short-term load and photovoltaic power forecasting[J]. Proceedings of the CSEE, 2019, 39 (8): 2221- 2230.
25	韩芳, 王学春, 靳宗信. 基于ACF的基音检测改进算法[J]. 现代电子技术, 2017, 40 (19): 71- 74, 78.
	HAN Fang, WANG Xuechun, JIN Zongxin. Improved pitch detection algorithm based on ACF[J]. Modern Electronics Technique, 2017, 40 (19): 71- 74, 78.

[1]	孙庆超, 李嘉靓, 江万里, 王若愚, 李植鹏, 胡亚荣, 朱健斌. 基于数据驱动时空网络的城市中长期电力负荷预测[J]. 中国电力, 2025, 58(3): 168-174.
[2]	王宇飞, 杜桐, 边伟国, 张钊, 刘慧婷, 杨丽君. 基于DTW K-medoids与VMD-多分支神经网络的多用户短期负荷预测[J]. 中国电力, 2024, 57(6): 121-130.
[3]	吴军英, 路欣, 刘宏, 张彬, 柴守亮, 刘蕴春, 王佳楠. 基于Spearman-GCN-GRU模型的超短期多区域电力负荷预测[J]. 中国电力, 2024, 57(6): 131-140.
[4]	陈中飞, 赵越, 蔡秋娜, 张乔榆, 王泽林, 戴晓娟, 陈雨果. 基于净负荷预测误差统计的电力系统爬坡能力充裕度评估[J]. 中国电力, 2024, 57(5): 50-60.
[5]	刘航, 申皓, 杨勇, 纪陵, 余洋. 基于高阶马尔可夫链的纯电重卡集群负荷预测[J]. 中国电力, 2024, 57(5): 61-69.
[6]	周颖, 白雪峰, 王阳, 邱敏, 孙冲, 武亚杰, 李彬. 面向虚拟电厂运营的温度敏感负荷分析与演变趋势研判[J]. 中国电力, 2024, 57(1): 9-17.
[7]	唐旭辰, 潮铸, 段秦尉, 苏炳洪, 陈卉灿. 基于分层测量数据的高压变电站概率负荷预测方法[J]. 中国电力, 2023, 56(8): 143-150.
[8]	贾巍, 黄裕春. 基于小样本数据差分扩容的微电网负荷预测方法[J]. 中国电力, 2023, 56(8): 151-156,165.
[9]	彭生江, 杨德州, 孙传帅, 袁铁江, 刘永成. 基于氢负荷需求的氢能系统容量规划[J]. 中国电力, 2023, 56(7): 13-20,32.
[10]	杨喜行, 黄纯, 申亚涛, 胡念恩, 万子恒. 基于损失功率匹配的配电线路故障定位方法[J]. 中国电力, 2022, 55(8): 113-120.
[11]	李文武, 石强, 李丹, 胡群勇, 唐芸, 梅锦超. 基于VMD和PSO-SVR的短期电力负荷多阶段优化预测[J]. 中国电力, 2022, 55(8): 171-177.
[12]	徐宇颂, 邹山花, 卢先领. 基于特征选择和组合模型的短期电力负荷预测[J]. 中国电力, 2022, 55(7): 121-127.
[13]	李丹, 张远航, 李黄强, 童华敏, 王凌云. 基于精准气象数据的地区电网短期负荷智能预测系统设计[J]. 中国电力, 2022, 55(7): 128-133.
[14]	郑豪丰, 杨国华, 康文军, 刘志远, 刘世涛, 伍弘, 张鸿皓. 基于多负荷特征和TCN-GRU神经网络的负荷预测[J]. 中国电力, 2022, 55(11): 142-148.
[15]	樊江川, 于昊正, 刘慧婷, 杨丽君, 安佳坤. 基于多分支门控残差卷积神经网络的短期电力负荷预测[J]. 中国电力, 2022, 55(11): 155-162,174.

考虑动态时间锚点和典型特征约束的年日均负荷曲线预测

Annual Daily Average Load Curve Prediction Considering Dynamic Time Anchors and Typical Feature Constraints

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 25

相关文章 15

编辑推荐

Metrics

模态框（Modal）标题

考虑动态时间锚点和典型特征约束的年日均负荷曲线预测

Annual Daily Average Load Curve Prediction Considering Dynamic Time Anchors and Typical Feature Constraints

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 25

相关文章 15

编辑推荐

Metrics