孟凡斌,南钰,武亚非,赵灵昊,卢长坤,乔金朋
MENG Fanbin,NAN Yu,WU Yafei,ZHAO Linghao,LU Changkun,QIAO Jinpeng

• 1:国网河南省电力公司开封供电公司
• 2:上海电力大学电气工程学院

摘要(Abstract)：

针对风力发电场景数据高维复杂的问题，提出一种基于谱归一化生成对抗网络与谱聚类的风力发电场景生成方法。首先，对生成器和判别器的两个深度神经网络进行对抗训练，并对判别器的卷积层进行谱归一化，通过增强模型的Lipschitz连续性约束来提高风电场景数据训练的稳定性和场景生成的质量。其次，采用基于改进高斯核函数的谱聚类方法提取风电特征，对高维数据进行降维，将生成的场景转化为典型风电场景集。最后，采用公开的WIND风电数据集进行仿真。仿真结果表明：所提方法能够显著降低生成样本的均方误差，精确捕捉风力发电的时空相关性；同时，基于高斯核函数的谱聚类方法能够有效对样本空间聚类。
To address the high-dimensional complexity of wind farm scenario data, a generation method based on spectral normalization generative adversarial networks and spectral clustering is proposed. First, adversarial training is performed on two deep neural networks of the generator and discriminator, with spectral normalization applied to the convolutional layers of the discriminator to enhance the Lipschitz continuity constraint, thereby improving the stability of data training and the quality of generated wind farm scenarios. Next, an improved Gaussian kernel-based spectral clustering method is used to extract wind power features and reduce the dimensionality of the data, transforming the generated scenarios into a set of typical wind farm scenarios. Finally, simulations are conducted using the publicly available WIND dataset. The simulation results indicate that the proposed method significantly reduces the mean squared errors of generated samples, accurately capturing the spatiotemporal correlations of wind power generation; the spectral clustering based on Gaussian kernel function effectively clusters the sample space.

关键词(KeyWords)： 电力系统;生成对抗网络;场景生成;谱归一化;谱聚类
power system;generative adversarial network;scenario generation;spectral normalization;spectral clustering

Abstract:

Keywords:

基金项目(Foundation): 上海市自然科学基金项目（22ZR1425500）;; 国网河南省电力公司科技项目（521790240005）

作者(Author): 孟凡斌,南钰,武亚非,赵灵昊,卢长坤,乔金朋
MENG Fanbin,NAN Yu,WU Yafei,ZHAO Linghao,LU Changkun,QIAO Jinpeng

DOI: 10.19585/j.zjdl.202412009

参考文献(References)：

• [1] DOLATABADI A, MOHAMMADI-IVATLOO B,ABAPOUR M,et al.Optimal stochastic design of wind integrated energy hub[J]. IEEE Transactions on Industrial Informatics,2017,13(5):2379-2388.
• [2]卢毓东，陈益.“双碳”目标下绿色人工智能技术研究综述[J].浙江电力，2023,42(10):45-56.LU Yudong,CHEN Yi.A review of green AI research under carbon peaking and neutrality goals[J].Zhejiang Electric Power,2023,42(10):45-56.
• [3]要金铭，赵书强，韦子瑜，等.基于场景分析的电力系统日前调度及其快速求解方法[J].电力自动化设备，2022,42(9):102-110.YAO Jinming,ZHAO Shuqiang,WEI Ziyu,et al. Dayahead dispatch and its fast solution method of power system based on scenario analysis[J].Electric Power Automation Equipment,2022,42(9):102-110.
• [4]艾欣，周树鹏，赵阅群.基于场景分析的含可中断负荷的优化调度模型研究[J].中国电机工程学报，2014,34（增刊1）:25-31.AI Xin,ZHOU Shupeng,ZHAO Yuequn.Research on optimal dispatch model considering interruptible loads based on scenario analysis[J].Proceedings of the CSEE,2014,34(S1):25-31.
• [5]苏向敬，张传坤，符杨，等.考虑多场景运行的不平衡主动配网电池储能系统两阶段优化配置[J].电力系统保护与控制，2023,51(10):88-97.SU Xiangjing,ZHANG Chuankun,FU Yang,et al.Twostage optimal placement of BESS in an unbalanced active distribution network considering multi-scenario operation[J].Power System Protection and Control,2023,51(10):88-97.
• [6]米阳，申杰，卢长坤，等.考虑含储能的三端智能软开关与需求侧响应的主动配电网有功无功协调优化[J].电力系统保护与控制，2024,52(3):104-118.MI Yang,SHEN Jie,LU Changkun,et al.Active and reactive power coordination optimization of an active distribution network considering a three-terminal soft open point with energy storage and demand response[J].Power System Protection and Control,2024,(3):104-118.
• [7]雷宇，杨明，韩学山.基于场景分析的含可再生能源系统机组组合的两阶段随机优化[J].电力系统保护与控制，2012,40(23):58-67.LEI Yu,YANG Ming,HAN Xueshan. A two-stage stochastic optimization of unit commitment considering wind power based on scenario analysis[J].Power System Protection and Control,2012,40(23):58-67.
• [8]贾梦瑶，王玉玮，宋明浩.基于时间生成对抗网络的风电随机场景预测[J].智慧电力，2023,51(8):59-66.JIA Mengyao,WANG Yuwei,SONG Minghao.Stochastic scenario prediction of wind power based on time generative adversarial networks[J].Smart Power,2023,51(8):59-66.
• [9]黄文琦，方必武，戴珍，等.基于多源数据图表示学习的风电出力预测方法[J].电力建设，2023,44(11):43-53.HUANG Wenqi,FANG Biwu,DAI Zhen,et al. Wind power output forecasting based on multi-source data graph representation learning[J]. Electric Power Construction,2023,44(11):43-53.
• [10] LEE D,BALDICK R.Load and wind power scenario generation through the generalized dynamic factor model[J].IEEE Transactions on Power Systems,2017,32(1):400-410.
• [11] FEIJOO A,VILLANUEVA D.Wind farm power distribution function considering wake effects[J].IEEE Transactions on Power Systems,2017,32(4):3313-3314.
• [12] H?YLAND K,KAUT M,WALLACE S W.A heuristic for moment-matching scenario generation[J]. Computational optimization and applications,2003,24(2):169-185.
• [13]朱瑞金，廖文龙，王玥珑，等.基于生成矩匹配网络的光伏和风电随机场景生成[J].高电压技术，2022,48(1):374-384.ZHU Ruijin,LIAO Wenlong,WANG Yuelong,et al.Stochastic scenarios generation for wind power and photovoltaic system based on generative moment matching network[J].High Voltage Engineering,2022,48(1):374-384.
• [14] CHEN Y Z,WANG Y S,KIRSCHEN D,et al.Modelfree renewable scenario generation using generative adversarial networks[J]. IEEE Transactions on Power Systems,2018,33(3):3265-3275.
• [15]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 network based monthly scenario analysis method for wind and photovoltaic power[J].Proceedings of the CSEE,2022,42(2):537-547.
• [16]杨秀，焦楷丹，孙改平，等.考虑负荷多无功用电场景的城市配电网无功优化配置[J].电力建设，2022,43(8):42-52.YANG Xiu,JIAO Kaidan,SUN Gaiping,et al. Reactive power optimization of urban distribution network considering multiple reactive power scenarios of loads[J].Electric Power Construction,2022,43(8):42-52.
• [17]袁铁江，杨洋，董力通.与典型日负荷匹配的风电出力场景构建方法[J].电力建设，2022,43(11):132-141.YUAN Tiejiang,YANG Yang,DONG Litong.Construction method of wind power output scenario matching with typical daily load[J].Electric Power Construction,2022,43(11):132-141.
• [18]徐胜蓝，司曹明哲，万灿，等.考虑双尺度相似性的负荷曲线集成谱聚类算法[J].电力系统自动化，2020,44(22):152-160.XU Shenglan,SI Caomingzhe,WAN Can,et al.Ensemble spectral clustering algorithm for load profiles considering dual-scale similarities[J]. Automation of Electric Power Systems,2020,44(22):152-160.
• [19] TAKERU MIYATO, TOSHIKI KATAOKA,MASANORI KOYAMA,et al.Spectral normalization for generative adversarial networks[J]. arXiv Preprint:1802.05957,2018.
• [20] DRAXL C,CLIFTON A,HODGE B M,et al.The wind integration national dataset(WIND)toolkit[J]. Applied Energy,2015,151:355-366.
• [21] VILLANUEVA D,FEIJOO A,PAZOS J L.Simulation of correlated wind speed data for economic dispatch evaluation[J].IEEE Transactions on Sustainable Energy,2012,3(1):142-149.
• [22] KIM M,RAMAKRISHNA R S.New indices for cluster validity assessment[J].Pattern Recognition Letters,2005,26(15):2353-2363.