高斯分布引导下负荷8760曲线全景最优化预测

doi:10.11930/j.issn.1004-9649.202411079

摘要/Abstract

摘要：

精确的长期负荷预测为电力系统规划提供良好的数据基础。现有长期负荷预测大多采用年、月等形式，时间分辨率较粗，难以反应负荷波动性。将电力负荷视作随机变量，提出一种高斯分布引导下负荷8760曲线全景最优化预测方法。首先，提出一种负荷8760曲线的长期负荷预测框架，该框架以全景式、精细化的结构描述负荷在未来整年即8760 h的长期变化，根据统计学特性分解为长期趋势、中期波动及短期变化，分析所包含的多时间尺度负荷指标。然后，以负荷取值概率最大为目标函数建立负荷概率优化模型，将全景负荷预测转化为优化问题，不同时间尺度负荷特性形成约束条件，求解优化模型并还原目标年负荷8760曲线。中国中部、南部两省实际负荷数据算例结果表明，所提方法通过确定负荷概率分布，以负荷8760曲线的形式准确描述全景负荷，有效提高长期预测精度，具备良好的泛化性与可解释性。

关键词: 负荷8760曲线, 全景负荷预测, 概率优化

Abstract:

Accurate long-term load forecasting provides base data for power system planning. However, most of the existing long-term load forecasting adopt the form of long time scale such as year and month, without considering the randomness and volatility of load. Taking the power load as variables, a panoramic optimization prediction method of 8760 load curve guided by Gaussian distribution is proposed. Firstly, a long-term panoramic load forecasting framework of load 8760 curve is proposed, which describes the long-term load changes in a panoramic and refined structure for a year which consist of 8760 h. The multi-time scale load indicators contained in the curve were analyzed and decomposed into long-term trend, medium-term fluctuation and short-term change according to statistical characteristics. The probability distribution of the maximum daily load with high fluctuation is determined by Gaussian hypothesis test. For other indicators with obvious statistical characteristics, the change rule is found on its time scale. Then, the load probability optimization model is established with the maximum load probability as the objective function, and the panoramic load prediction is transformed into an optimization problem. By predicting the load characteristics of different time scales, the constraints of the model are formed, and the optimization model is solved and the target annual load 8760 curve is restored. The results of the actual load data arithmetic examples in the central and southern provinces of China show that the proposed method accurately describes the panoramic loads in the form of load 8760 curves by determining the load probability distributions, which can effectively improve the accuracy of long-term forecasting and has great generalizability and interpretability.

Key words: 8760 load curve, panoramic load forecasting, probabilistic optimization

罗超, 倪恬, 陈凌云, 康义, 侯慧, 吴细秀. 高斯分布引导下负荷8760曲线全景最优化预测[J]. 中国电力, 2025, 58(8): 31-40.

LUO Chao, NI Tian, CHEN Lingyun, KANG Yi, HOU Hui, WU Xixiu. Panoramic Optimal Prediction of Load 8760 Curve Guided by Gaussian Distribution[J]. Electric Power, 2025, 58(8): 31-40.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202411079

https://www.electricpower.com.cn/CN/Y2025/V58/I8/31

图/表 6

参考文献 26

1	孙庆超, 李嘉靓, 江万里, 等. 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. DOI
	SUN Qingchao, LI Jialiang, JIANG Wanli, et al. Mid-long term urban power load forecasting based on data-driven spatio-temporal networks[J]. Electric Power, 2025, 58 (3): 168- 174. DOI
2	李丹, 贺帅, 颜伟, 等. 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. DOI
	LI Dan, HE Shuai, YAN Wei, et al. Annual daily average load curve prediction considering dynamic time anchors and typical feature constraints[J]. Electric Power, 2024, 57 (11): 36- 47. DOI
3	曲晓黎, 尤琦, 李文晴, 等. 气象因子相对危险度在电网用电负荷预测中的应用[J]. 南方能源建设, 2024, 11 (1): 166- 175.
	QU Xiaoli, YOU Qi, LI Wenqing, et al. Application of relative risk of meteorological factors in power grid electricity load forecasting[J]. Southern Energy Construction, 2024, 11 (1): 166- 175.
4	张明泽, 栾文鹏, 艾欣, 等. 基于边缘计算的台区短期负荷预测方法[J]. 电测与仪表, 2024, 61 (4): 93- 99.
	ZHANG Mingze, LUAN Wenpeng, AI Xin, et al. Short-term substation load forecasting method based on edge computing[J]. Electrical Measurement & Instrumentation, 2024, 61 (4): 93- 99.
5	杨坚, 赵洁, 汤义勤, 等. 基于序列分解的母线负荷降噪预测方法[J]. 浙江电力, 2023, 42 (12): 81- 87.
	YANG Jian, ZHAO Jie, TANG Yiqin, et al. A noise reduction forecasting method of bus load based on sequence decomposition[J]. Zhejiang Electric Power, 2023, 42 (12): 81- 87.
6	RUBASINGHE O, ZHANG X N, CHAU T K, et al. A novel sequence to sequence data modelling based CNN-LSTM algorithm for three years ahead monthly peak load forecasting[J]. IEEE Transactions on Power Systems, 2024, 39 (1): 1932- 1947. DOI
7	BASHIRI BEHMIRI N, FEZZI C, RAVAZZOLO F. Incorporating air temperature into mid-term electricity load forecasting models using time-series regressions and neural networks[J]. Energy, 2023, 278, 127831. DOI
8	史笑宇, 王欣, 王钢. 基于PSO-SVM组合模型的短期电力需求预测[J]. 东北电力大学学报, 2024, 44 (6): 74- 81.
	SHI Xiaoyu, WANG Xin, WANG Gang. Short-term power demand forecasting based on PSO-SVM combined model[J]. Journal of Northeast Electric Power University, 2024, 44 (6): 74- 81.
9	LUO J, HONG T, GAO Z M, et al. A robust support vector regression model for electric load forecasting[J]. International Journal of Forecasting, 2023, 39 (2): 1005- 1020. DOI
10	XIE Y L, JIN M P, ZOU Z P, et al. Real-time prediction of docker container resource load based on a hybrid model of ARIMA and triple exponential smoothing[J]. IEEE Transactions on Cloud Computing, 2022, 10 (2): 1386- 1401. DOI
11	TANG T, JIANG W H, ZHANG H, et al. GM(1, 1) based improved seasonal index model for monthly electricity consumption forecasting[J]. Energy, 2022, 252, 124041. DOI
12	赵会茹, 赵一航, 郭森. 基于互补集合经验模态分解和长短期记忆神经网络的短期电力负荷预测[J]. 中国电力, 2020, 53 (6): 48- 55. DOI
	ZHAO Huiru, ZHAO Yihang, GUO Sen. Short-term load forecasting based on complementary ensemble empirical mode decomposition and long short-term memory[J]. Electric Power, 2020, 53 (6): 48- 55. DOI
13	吴军英, 路欣, 刘宏, 等. 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. DOI
	WU Junying, LU Xin, LIU Hong, et al. Ultra-short-term multi-region power load forecasting based on spearman-GCN-GRU model[J]. Electric Power, 2024, 57 (6): 131- 140. DOI
14	刘杰, 从兰美, 夏远洋, 等. 基于DBO-VMD和IWOA-BILSTM神经网络组合模型的短期电力负荷预测[J]. 电力系统保护与控制, 2024, 52 (8): 123- 133.
	LIU Jie, CONG Lanmei, XIA Yuanyang, et al. Short-term power load prediction based on DBO-VMD and an IWOA-BILSTM neural network combination model[J]. Power System Protection and Control, 2024, 52 (8): 123- 133.
15	HE X, ZHAO W L, GAO Z J, et al. A hybrid prediction interval model for short-term electric load forecast using Holt-Winters and Gate Recurrent Unit[J]. Sustainable Energy, Grids and Networks, 2024, 38, 101343. DOI
16	DONG H J, ZHU J Z, LI S L, et al. Short-term residential household reactive power forecasting considering active power demand via deep Transformer sequence-to-sequence networks[J]. Applied Energy, 2023, 329, 120281. DOI
17	HUANG X S, WU D, BOULET B. MetaProbformer for charging load probabilistic forecasting of electric vehicle charging stations[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24 (10): 10445- 10455. DOI
18	王炜, 冯斌, 黄刚, 等. 基于自注意力编码器和深度神经网络的短期净负荷预测[J]. 中国电机工程学报, 2023, 43 (23): 9072- 9084.
	WANG Wei, FENG Bin, HUANG Gang, et al. Short-term net load forecasting based on self-attention encoder and deep neural network[J]. Proceedings of the CSEE, 2023, 43 (23): 9072- 9084.
19	罗澍忻, 麻敏华, 蒋林, 等. 考虑多时间尺度数据的中长期负荷预测方法[J]. 中国电机工程学报, 2020, 40 (S1): 11- 19.
	LUO Shuxin, MA Minhua, JIANG Lin, et al. Medium and long-term load forecasting method considering multi-time scale data[J]. Proceedings of the CSEE, 2020, 40 (S1): 11- 19.
20	康重庆, 夏清, 张伯明. 电力系统负荷预测研究综述与发展方向的探讨[J]. 电力系统自动化, 2004, 28 (17): 1- 11. DOI
	KANG Chongqing, XIA Qing, ZHANG Boming. Review of power system load forecasting and its development[J]. Automation of Electric Power Systems, 2004, 28 (17): 1- 11. DOI
21	STAGGE J H, TALLAKSEN L M, GUDMUNDSSON L, et al. Candidate distributions for climatological drought indices (SPI and SPEI)[J]. International Journal of Climatology, 2015, 35 (13): 4027- 4040. DOI
22	JIANG Z X, ZHANG L L, JI T Y. NSDAR: a neural network-based model for similar day screening and electric load forecasting[J]. Applied Energy, 2023, 349, 121647. DOI
23	KHAN W, SOMERS W, WALKER S, et al. Comparison of electric vehicle load forecasting across different spatial levels with incorporated uncertainty estimation[J]. Energy, 2023, 283, 129213. DOI
24	唐旭辰, 潮铸, 段秦尉, 等. 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. DOI
	TANG Xuchen, CHAO Zhu, DUAN Qinwei, et al. Probabilistic load forecasting method of high voltage substation based on hierarchical measurement data[J]. Electric Power, 2023, 56 (8): 143- 150. DOI
25	卢皓天, 刘少鹏, 王凯. 基于多目标粒子群算法的光伏制氢系统容量优化方法[J]. 南方能源建设, 2025, 12 (3): 133- 143.
	LU Haotian, LIU Shaopeng, WANG Kai. Capacity optimization method for photovoltaic hydrogen production systems based on multi-objective particle swarm algorithm[J]. Southern Energy Construction, 2025, 12 (3): 133- 143.
26	刘子祺, 苏婷婷, 何佳阳, 等. 基于多目标粒子群算法的配电网储能优化配置研究[J]. 综合智慧能源, 2023, 45 (6): 9- 16.
	LIU Ziqi, SU Tingting, HE Jiayang, et al. Research on the optimal allocation of energy storage in distribution network based on multi-objective particle swarm optimization algorithm[J]. Integrated Intelligent Energy, 2023, 45 (6): 9- 16.

类别	模型	简称
基准模型	2 021年实际负荷8760曲线	M1
特征映射模型	RF	M2
	SVM	M3
	KNN	M4
自相关回归模型	LSTM	M5
	GRU	M6
	Transformer	M7
概率优化模型	本文所提模型	M8

类别	模型	简称
基准模型	2 021年实际负荷8760曲线	M1
特征映射模型	RF	M2
	SVM	M3
	KNN	M4
自相关回归模型	LSTM	M5
	GRU	M6
	Transformer	M7
概率优化模型	本文所提模型	M8

模型	算例	超参数	寻优结果
M2	1	决策树数量	96
		最大深度	12
		划分最小样本数	21
		叶节点最小样本数	1
	2	决策树数量	148
		最大深度	100
		划分最小样本数	2
		叶节点最小样本数	89
M3	1	惩罚参数	200
	1	损失函数宽度	0.1
	2	惩罚参数	200
	2	损失函数宽度	0.1
M4	1	最近邻数量	61
M4	2	最近邻数量	100
M5	1	神经元数量	64/30
	1	时间步长	24
	2	神经元数量	55/21
	2	时间步长	18
M6	1	神经元数量	100/53
	1	时间步长	7
	2	神经元数量	98/67
	2	时间步长	14
M7	1	维度	200
	1	头数	8
	2	维度	134
	2	头数	8

模型	算例	超参数	寻优结果
M2	1	决策树数量	96
		最大深度	12
		划分最小样本数	21
		叶节点最小样本数	1
	2	决策树数量	148
		最大深度	100
		划分最小样本数	2
		叶节点最小样本数	89
M3	1	惩罚参数	200
	1	损失函数宽度	0.1
	2	惩罚参数	200
	2	损失函数宽度	0.1
M4	1	最近邻数量	61
M4	2	最近邻数量	100
M5	1	神经元数量	64/30
	1	时间步长	24
	2	神经元数量	55/21
	2	时间步长	18
M6	1	神经元数量	100/53
	1	时间步长	7
	2	神经元数量	98/67
	2	时间步长	14
M7	1	维度	200
	1	头数	8
	2	维度	134
	2	头数	8