基于CGAN台风气象下负荷场景生成

doi:10.11930/j.issn.1004-9649.202405076

中国电力 ›› 2025, Vol. 58 ›› Issue (2): 176-185.DOI: 10.11930/j.issn.1004-9649.202405076

基于CGAN台风气象下负荷场景生成

罗萍萍¹(), 盛奥¹(), 林济铿²(), 王忠岳³, 李启本⁴, 周平⁴

1. 上海电力大学电气工程学院，上海　200090
2. 同济大学电子与信息工程学院，上海　201804
3. 国网上海市电力公司市区供电公司，上海　200080
4. 国网上海市电力公司松江供电公司，上海　201600

收稿日期:2024-05-17 接受日期:2024-11-22 出版日期:2025-02-28 发布日期:2025-02-25
作者简介:罗萍萍（1969—），女，副教授，从事电力系统继电保护研究，E-mail：luopingping@shiep.edu.cn
盛奥（1997—），男，硕士研究生，从事源荷数据挖掘研究，E-mail：1772876090@qq.com
林济铿（1967—），男，通信作者，教授，从事电力系统优化调度、EMS/DEMS、人工智能、电力市场等研究，E-mail：mejklin@126.com
基金资助:
国家自然科学基金资助项目（51177107）。

CGAN-Based Load Scenario Generation under Typhoon Weather

Pingping LUO¹(), Ao SHENG¹(), Jikeng LIN²(), Zhongyue WANG³, Qiben LI⁴, Ping ZHOU⁴

1. School of Electric Power Engineering, Shanghai University of Electric Power, Shanghai 200090, China
2. Electronic and Information Engineering, Tongji University, Shanghai 201804, China
3. State Grid Shanghai Electric Power Company Urban Power Supply Company, Shanghai 200080, China
4. State Grid Songjiang Power Supply Company, SMEPC, Shanghai 201600, China

Received:2024-05-17 Accepted:2024-11-22 Online:2025-02-28 Published:2025-02-25
Supported by:
This work is supported by National Natural Science Foundation of China (No.51177107).

摘要/Abstract

摘要：

台风天气下负荷水平的剧烈波动威胁电网平衡，相应地台风气象条件下的负荷场景生成受到电力公司的重视。因此，提出一种面向台风天气基于条件生成对抗网络（conditional generative adversarial network，CGAN）模型的负荷场景生成算法。首先，针对台风样本登陆位置分散、持续周期不同及等级不同的特点，提出一种台风下负荷样本的拆分及标签给定方法。然后，针对台风气象下负荷样本数量稀少，提出一种基于条件概率的样本扩充策略以扩充样本集。最后，为了进一步提升样本集的有效性，基于迁徙训练思想，先用正常气象下的负荷样本对CGAN进行训练，然后再采用台风样本集训练CGAN，在模型训练完成后，输入随机噪声与台风标签即可生成对应的负荷场景。算例证实了所提模型及算法的有效性和先进性。

关键词: 台风气象, 负荷场景生成, 数据扩充, 标签, 人工智能

Abstract:

The violent fluctuation of power load level under typhoon weather threatens the power balance of power grid. Therefore, load scenario generation under typhoon weather conditions has attracted increasing attention from power supply companies. A load scenario generation algorithm based on conditional generative adversarial network (CGAN) model for typhoon weather is proposed. Firstly, considering the fact that the typhoon samples have the characteristics of scattered landing locations, different duration periods and different grades, a load sample classification and label setting method for typhoon weather is proposed. Then, a sample expansion strategy based on conditional probability is proposed to expand the sample set to solve the problem of scarce load samples under typhoon weather. Finally, in order to further improve the actual effectiveness of the sample set, based on the idea of migration training, the load samples under normal weather are firstly used to train the CGAN, and then the typhoon sample sets are applied to train CGAN. After the model training is completed, the corresponding load scenarios can be quickly generated by inputting random noise and typhoon labels. The effectiveness and advancement of the proposed model and algorithm are verified by data set from a practical power system.

Key words: typhoon weather, load scenario generation, data expansion, label setting, artificial intelligence

罗萍萍, 盛奥, 林济铿, 王忠岳, 李启本, 周平. 基于CGAN台风气象下负荷场景生成[J]. 中国电力, 2025, 58(2): 176-185.

Pingping LUO, Ao SHENG, Jikeng LIN, Zhongyue WANG, Qiben LI, Ping ZHOU. CGAN-Based Load Scenario Generation under Typhoon Weather[J]. Electric Power, 2025, 58(2): 176-185.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202405076

https://www.electricpower.com.cn/CN/Y2025/V58/I2/176

图/表 10

图 1 类型为$ {T^{\text{y}}} $持续$ {C^{\text{y}}} $日的台风下负荷标签给定

Fig.1 Load label setting under a typhoon of type $ {T^{\text{y}}} $ for $ {C^{\text{y}}} $ days

图 2 CGAN结构

Fig.2 CGAN structure

表 1 生成器网络结构

Table 1 Structure of the generator

结构	神经元个数	动态均值动量	激活函数
输入层	89	/	0.2
隐藏层1	64	0.8	0.2
隐藏层2	128	0.8	0.2
输出层	96	/	—

表 2 判别器网络结构

Table 2 Structure of the discriminator

结构	神经元个数	动态均值动量	激活函数
输入层	123	/	0.2
隐藏层1	128	0.8	0.2
隐藏层2	64	0.8	0.2
输出层	1	/	—

图 3 基于CGAN的台风气象下负荷场景生成示意

Fig.3 Flow chart of CGAN-based load scenario generation under typhoon weather

图 4 判别器损失函数变化曲线

Fig.4 Variation curve of the discriminator loss function

图 5 不同训练方法生成样本质量模拟

Fig.5 Sample quality generated simulation by different training methods

图 6 生成场景自相关系数与偏自相关系数箱型

Fig.6 Box diagram of autocorrelation coefficient and partial autocorrelation coefficient of scenario generation

图 7 不同方法训练模型生成结果置信度结果对比

Fig.7 Confidence comparison of generation results by different training methods

表 3 生成负荷场景指标计算结果

Table 3 Calculation results of load generation scenario indicators

置信度/%	台风类型			对比方法		未扩充样本训练		扩充样本训练
置信度/%	登录位置	等级	持续天数	覆盖率/%	负荷区间	覆盖率/%	负荷区间	覆盖率/%	负荷区间
90	3	3	4	87.5	0.63	91.1	0.51	92.9	0.49
80				66.1	0.46	69.8	0.36	71.8	0.34
50				42.5	0.30	50.2	0.25	51.8	0.21
90	1	2	6	88.2	0.65	89.3	0.57	90.1	0.56
80				65.1	0.47	68.6	0.44	70.4	0.42
50				41.6	0.32	43.2	0.29	44.7	0.26

参考文献 27

1	李明节, 梁志峰, 许涛, 等. 基于敏感气温空间分布的度夏度冬日最大负荷预测与应用研究[J]. 电网技术, 2023, 47 (3): 1088- 1097.
	LI Mingjie, LIANG Zhifeng, XU Tao, et al. Prediction and application of maximum daily load in summer and winter based on spatial distribution of sensitive temperatures[J]. Power System Technology, 2023, 47 (3): 1088- 1097.
2	樊晓伟, 王瑞妙, 杨海峰, 等. 计及源荷不确定的综合能源微电网集群优化运行[J]. 电力建设, 2024, 45 (8): 128- 139.
	FAN Xiaowei, WANG Ruimiao, YANG Haifeng, et al. Optimization operation of integrated energy microgrid cluster considering source-load uncertainty[J]. Electric Power Construction, 2024, 45 (8): 128- 139.
3	朱海军, 鞠立伟, 杨慧, 等. 计及不确定性的新型灵活性资源两阶段鲁棒配置优化模型[J]. 电力建设, 2024, 45 (7): 1- 11.
	ZHU Haijun, JU Liwei, YANG Hui, et al. Optimal two-stage robust configuration model and algorithm of new flexibility resources considering uncertainty[J]. Electric Power Construction, 2024, 45 (7): 1- 11.
4	李滨, 陆明珍. 考虑实时气象耦合作用的地区电网短期负荷预测建模[J]. 电力系统自动化, 2020, 44 (17): 60- 68. DOI
	LI Bin, LU Mingzhen. Short-term load forecasting modeling of regional power grid considering real-time meteorological coupling effect[J]. Automation of Electric Power Systems, 2020, 44 (17): 60- 68. DOI
5	申洪涛, 李飞, 史轮, 等. 基于气象数据降维与混合深度学习的短期电力负荷预测[J]. 电力建设, 2024, 45 (1): 13- 21.
	SHEN Hongtao, LI Fei, SHI Lun, et al. Short-term power load forecasting based on reduction of meteorological data dimensionality and hybrid deep learning[J]. Electric Power Construction, 2024, 45 (1): 13- 21.
6	李雪玲, 刘洋, 李振伟, 等. 基于气象分型改进构造不确定集的多微网低碳鲁棒经济调度[J]. 电力建设, 2023, 44 (8): 142- 156.
	LI Xueling, LIU Yang, LI Zhenwei, et al. Robust low-carbon economic dispatch of multiple microgrids based on improved uncertainty set of meteorological classification[J]. Electric Power Construction, 2023, 44 (8): 142- 156.
7	李宝聚, 齐宏伟, 傅吉悦, 等. 极端气象天气对新能源运行影响分析[J]. 吉林电力, 2022, 50 (1): 10- 13. DOI
	LI Baoju, QI Hongwei, FU Jiyue, et al. Analysis on the impact of extreme weather on new energy operation[J]. Jilin Electric Power, 2022, 50 (1): 10- 13. DOI
8	李帅, 李哲, 梁允, 等. 气象因素对电力行业的影响研究[J]. 东北电力技术, 2019, 40 (8): 7- 10. DOI
	LI Shuai, LI Zhe, LIANG Yun, et al. Research on influence of meteorology on electric power industry[J]. Northeast Electric Power Technology, 2019, 40 (8): 7- 10. DOI
9	方华亮, 李大虎, 彭辉, 等. 基于“互联网+” 的分散式太阳能规划方法[J]. 中国电机工程学报, 2017, 37 (5): 1316- 1323.
	FANG Hualiang, LI Dahu, PENG Hui, et al. Distributed solar energy planning method based on Internet plus[J]. Proceedings of the CSEE, 2017, 37 (5): 1316- 1323.
10	MOTLAGH O, BERRY A, O’NEIL L. Clustering of residential electricity customers using load time series[J]. Applied Energy, 2019, 237, 11- 24. DOI
11	李博, 孙建军, 余攀, 等. 基于负荷聚类与网络等效的配电网多维典型场景生成方法[J]. 中国电机工程学报, 2021, 41 (8): 2661- 2670.
	LI Bo, SUN Jianjun, YU Pan, et al. A multi-dimensional typical scenarios generation algorithm for distribution network based on load clustering and network structure equivalence[J]. Proceedings of the CSEE, 2021, 41 (8): 2661- 2670.
12	曲凯, 李湃, 黄越辉, 等. 面向新能源消纳能力评估的年负荷序列建模及场景生成方法[J]. 电力系统自动化, 2021, 45 (1): 123- 131.
	QU Kai, LI Pai, HUANG Yuehui, et al. Modeling and scenario generation method of annual load series for evaluation of renewable energy accommodation capacity[J]. Automation of Electric Power Systems, 2021, 45 (1): 123- 131.
13	李辉, 任洲洋, 胡博, 等. 基于时序生成对抗网络的月度风光发电功率场景分析方法[J]. 中国电机工程学报, 2022, 42 (2): 537- 547.
	LI Hui, REN Zhouyang, HU Bo, et al. A sequential generative adversarial network based monthly scenario analysis method for wind and photovoltaic power[J]. Proceedings of the CSEE, 2022, 42 (2): 537- 547.
14	董骁翀, 孙英云, 蒲天骄. 基于条件生成对抗网络的可再生能源日前场景生成方法[J]. 中国电机工程学报, 2020, 40 (17): 5527- 5535.
	DONG Xiaochong, SUN Yingyun, PU Tianjiao. Day-ahead scenario generation of renewable energy based on conditional GAN[J]. Proceedings of the CSEE, 2020, 40 (17): 5527- 5535.
15	SARMAS E, SPILIOTIS E, STAMATOPOULOS E, et al. Short-term photovoltaic power forecasting using meta-learning and numerical weather prediction independent long short-term memory models[J]. Renewable Energy, 2023, 216, 118997. DOI
16	肖白, 黄钰茹, 姜卓, 等. 数据匮乏场景下采用生成对抗网络的空间负荷预测方法[J]. 中国电机工程学报, 2020, 40 (24): 7990- 8001.
	XIAO Bai, HUANG Yuru, JIANG Zhuo, et al. The method of spatial load forecasting based on the generative adversarial network for data scarcity scenarios[J]. Proceedings of the CSEE, 2020, 40 (24): 7990- 8001.
17	JING H, ZHAO C H. A data expansion based piecewise regression strategy for incrementally monitoring the wind turbine with power curve[J]. Journal of Central South University, 2023, 30 (5): 1601- 1617. DOI
18	PENG X S, YANG Z M, LI Y H, et al. Short-term wind power prediction based on stacked denoised auto-encoder deep learning and multi-level transfer learning[J]. Wind Energy, 2023, 26 (10): 1066- 1081. DOI
19	朱涛, 陈翔, 高强, 等. 基于Logsitic回归模型和自适应密度聚类算法的分行业负荷增长规律[J]. 电网与清洁能源, 2019, 35 (5): 20- 28. DOI
	ZHU Tao, CHEN Xiang, GAO Qiang, et al. Law of load growth in different industries based on logsitic regression model and adaptive density clustering algorithm[J]. Power System and Clean Energy, 2019, 35 (5): 20- 28. DOI
20	郑凌蔚, 刘士荣, 周文君, 等. 并网型可再生能源发电系统容量配置与优化[J]. 电力系统保护与控制, 2014, 42 (17): 31- 37. DOI
	ZHENG Lingwei, LIU Shirong, ZHOU Wenjun, et al. Capacity configuration and optimization of grid-connected renewable energy power generation system[J]. Power System Protection and Control, 2014, 42 (17): 31- 37. DOI
21	陈国民, 杨梦琪, 张喜平, 等. 2020年西北太平洋和南海台风预报精度评定[J]. 气象, 2022, 48 (4): 516- 525. DOI
	CHEN Guomin, YANG Mengqi, ZHANG Xiping, et al. Verification on forecasts of typhoons over western North Pacific and South China Sea in 2020[J]. Meteorological Monthly, 2022, 48 (4): 516- 525. DOI
22	欧阳誉波. 基于混合藤Copula模型的多风电场出力相关性建模及其在无功优化中的应用[D]. 成都: 西南交通大学, 2016.
	OUYANG Yubo. Modeling of multi-dimensional wind farms output correlation based on mixture vine Copula structures and its application in reactive power optimization[D]. Chengdu: Southwest Jiaotong University, 2016.
23	MIRZA M, OSINDERO S. Conditional generative adversarial nets[EB/OL].(2014-11-06) [2024-05-16]. https://arxiv.org/abs/1411.1784.
24	ARJOVSKY M, BOTTOU L. Towards principled methods for training generative adversarial networks[EB/OL]. (2017-01-17) [2024-05-16].http://arxiv.org/abs/1701.04862v1.
25	HOCHREITER R, PFLUG G C. Financial scenario generation for stochastic multi-stage decision processes as facility location problems[J]. Annals of Operations Research, 2007, 152 (1): 257- 272. DOI
26	ZHUANG F Z, QI Z Y, DUAN K Y, et al. A comprehensive survey on transfer learning[J]. Proceedings of the IEEE, 2021, 109 (1): 43- 76. DOI
27	李湃, 刘纯, 黄越辉, 等. 基于隐马尔科夫模型的多风电场相关性出力时间序列建模方法[J]. 中国电机工程学报, 2019, 39 (19): 5683- 5691.
	LI Pai, LIU Chun, HUANG Yuehui, et al. Modeling correlated power time series of multiple wind farms based on hidden Markov model[J]. Proceedings of the CSEE, 2019, 39 (19): 5683- 5691.

[1]	高志远, 庄卫金, 李峰, 于芳, 张鸿, 王艳, 夏旻. 调控领域人工智能应用的高复用性验证平台[J]. 中国电力, 2025, 58(3): 142-150.
[2]	薛溟枫, 毛晓波, 肖浩, 周毅斌, 浦骁威, 裴玮. 基于联邦学习的综合能源微网群协同优化运行方法[J]. 中国电力, 2023, 56(12): 164-173.
[3]	郝玲玲, 朱永利, 王永正. 基于DCAE-KSSELM的变压器故障诊断方法[J]. 中国电力, 2022, 55(2): 125-130.
[4]	刘文君, 董明, 徐元孚, 韩强, 王鑫, 许雷, 杜明. 电力设备运行状态大数据标签体系与关键技术[J]. 中国电力, 2022, 55(1): 126-132.
[5]	华志刚, 范佳卿, 郭荣, 汪勇, 吴潇翔. 人工智能技术在火电行业的应用探讨[J]. 中国电力, 2021, 54(7): 198-207.
[6]	赵永良, 付鑫, 郭阳, 边迎迎, 王思宁. 基于深度学习和图像识别的电力配件智能出入库[J]. 中国电力, 2021, 54(3): 55-60.
[7]	李博, 方彤. 北斗卫星导航系统（BDS）在智能电网的应用与展望[J]. 中国电力, 2020, 53(8): 107-116.
[8]	王利利, 张琳娟, 许长清, 武宏波, 李翼铭, 黄玉晶. 能源互联网背景下园区用户画像及成熟度评价模型研究[J]. 中国电力, 2020, 53(8): 19-28.
[9]	杨晓楠, 孙博, 郎燕生. 基于深度学习的特高压直流闭锁故障智能调度决策[J]. 中国电力, 2020, 53(6): 8-17.
[10]	郭敬东, 陈彬, 王仁书, 王佳宇, 仲林林. 基于YOLO的无人机电力线路杆塔巡检图像实时检测[J]. 中国电力, 2019, 52(7): 17-23.
[11]	罗昊, 苏盛, 杨浩, 林楠, 袁晨. 基于FPGA的电力巡线无人机硬件加密通信方法[J]. 中国电力, 2019, 52(7): 11-16.
[12]	岳国良, 路艳巧, 常浩, 孙翠英. 一种基于可见光巡检图像的杂草智能识别方法[J]. 中国电力, 2019, 52(11): 138-144,174.
[13]	张晋宾, 周四维. 智能电厂概念及体系架构模型研究[J]. 中国电力, 2018, 51(10): 2-7,42.
[14]	李博, 高志远. 人工智能技术在智能电网中的应用分析和展望[J]. 中国电力, 2017, 50(12): 136-140.
[15]	李海山, 雎刚, 毛晓飞, 余廷芳. 基于智能计算的锅炉燃烧优化指导系统及其应用[J]. 中国电力, 2014, 47(7): 1-5.

基于CGAN台风气象下负荷场景生成

CGAN-Based Load Scenario Generation under Typhoon Weather

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 27

相关文章 15

编辑推荐

Metrics

模态框（Modal）标题

基于CGAN台风气象下负荷场景生成

CGAN-Based Load Scenario Generation under Typhoon Weather

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 27

相关文章 15

编辑推荐

Metrics