| 188 | 1 | 175 |
| 下载次数 | 被引频次 | 阅读次数 |
为优化配电网网格空间的负荷聚类策略，提升负荷聚类的精准度和效率，提出一种基于改进K-means和聚类生成对抗网络（cluster generative adversarial network,Cluster-GAN）的配电网网格划分和负荷聚类方法，通过改进K-means算法对配电网负荷数据进行网格划分，接着采用聚类生成对抗网络对网格化负荷数据的精细化聚类。首先，通过引入改进的K-means算法，基于综合网格划分指标对配电网负荷数据进行初步处理。然后，利用融入聚类损失的聚类生成对抗网络模型，对网格内复杂的负荷数据进行深度聚类分析。Cluster-GAN通过生成包含类簇信息的样本数据，在潜在空间内实现高效聚类，有效规避了传统聚类方法易陷入局部最优的难题，显著提升了负荷聚类的准确性和效率。仿真结果显示，该方法能够精确描绘配电网的负荷分布特征，为新型电力系统下配电网及负荷的科学管理、精准刻画及能源优化调度奠定了坚实的数据基础与技术支撑。
Abstract：To optimize the load clustering strategy in distribution network grid space and improve the accuracy and efficiency of load clustering,this paper proposes a meshing and load clustering method in distribution network grid space based on improved K-means and cluster generative adversarial network(Cluster-GAN). The method performs preliminary meshing of distribution network load data by improved K-means algorithm,and then uses Cluster-GAN for refined clustering of gridded load data. Firstly,the initial processing of the distribution network load data is performed by introducing the improved K-means algorithm based on the integrated meshing metrics. Subsequently,a Cluster-GAN model incorporating clustering loss is used to perform deep clustering analysis of complex load data within the grid.By generating sample data rich in clustering information,Cluster-GAN achieves efficient clustering in potential space,which effectively circumvents the difficulty that traditional clustering methods are prone to fall into local optimum,and significantly improves the accuracy and efficiency of load clustering.Simulation results show that the method can accurately depict the load distribution characteristics of the distribution network,which lays a solid data foundation and technical support for the scientific management and accurate portrayal of the distribution network and loads under the new power system.
[1]徐嘉豪，汪泽原.新型电力系统及“双碳”下配电网规划技术及策略[J].中国设备工程，2024(17):203-205.XU Jiahao,WANG Zeyuan. New power system and distribution network planning technology and strategy under“double carbon”[J].China Plant Engineering,2024(17):203-205.
[2]郁海彬，袁欣诚，唐亮，等.新型电力系统下虚拟电厂参与电-碳联合市场的竞标策略与价格传导机制[J].电力大数据，2024,27(4):1-15.YU Haibin,YUAN Xincheng,TANG Liang,et al.Bidding strategy and price transmission mechanism of virtual power plants participating in the electricity carbon joint market under the new power system[J].Power Systems and Big Data,2024,27(4):1-15.
[3]吴鸣，张楠春，梁英，等.新型电力系统背景下微电网技术研究与发展[J].新型电力系统，2024,2(3):251-271.WU Ming,ZHANG Nanchun,LIANG Ying,et al. Research and development of microgrid technology in the context of new type power system[J].New Type Power Systems,2024,2(3):251-271.
[4]王博宇，张喻，郭昱成.可再生能源与电力系统的可持续发展[J].建材发展导向，2024,22(5):120-122.WANG Boyu,ZHANG Yu,GUO Yucheng.Renewable energy and sustainable development of power system[J]. Development Guide to Building Materials,2024,22(5):120-122.
[5]倪威中，何俊，熊凤龙.新型电力系统形势下基于负荷特性的负荷预测研究[J].电工技术，2022(18):140-144.NI Weizhong,HE Jun,XIONG Fenglong. Research on load forecasting based on load characteristics under the situation of new power system[J].Electric Engineering,2022(18):140-144.
[6]桂海涛，秦丽文，杨鑫，等.基于数据融合的高效可靠智能电网通信技术[J].科学技术与工程，2021,21(33):14173-14180.GUI Haitao,QIN Liwen,YANG Xin,et al. Efficient and reliable smart grid communication technology based on data fusion[J].Science Technology and Engineering,2021,21(33):14173-14180.
[7]王森，刘琛，邢帅杰.K-means聚类算法研究综述[J].华东交通大学学报，2022,39(5):119-126.WANG Sen,LIU Chen,XING Shuaijie. Review on K-means clustering algorithm[J].Journal of East China Jiaotong University,2022,39(5):119-126.
[8]张凯，冯剑，刘建华，等.基于模糊C均值聚类算法的用电行为模式分类[J].电力需求侧管理，2022,24(3):98-103.ZHANG Kai,FENG Jian,LIU Jianhua,et al. Power consumption behavior pattern classification based on fuzzy C-mean clustering algorithm[J]. Power Demand Side Management,2022,24(3):98-103.
[9]唐雅洁，阎洁，李玉浩，等.基于深度嵌入聚类的风光水典型联合出力场景提取[J].浙江电力，2023,42(4):36-44.TANG Yajie,YAN Jie,LI Yuhao,et al. Extraction of typical combined output scenarios of wind-solar-hydropower generation based on deep embedding clustering[J].Zhejiang Electric Power,2023,42(4):36-44.
[10]余浩，高镱滈，潘险险，等.基于改进高斯混合模型的变电站负荷聚类算法[J].全球能源互联网，2024,7(5):591-601.YU Hao,GAO Yihao,PAN Xianxian,et al. Substation load clustering algorithm based on improved Gaussian mixture model[J]. Journal of Global Energy Interconnection,2024,7(5):591-601.
[11] SUN Z X,ZHAO S S,ZHANG J X. Short-term wind power forecasting on multiple scales using VMD decomposition,K-means clustering and LSTM principal computing[J].IEEE Access,2019,7:166917-166929.
[12] LIU F,DONG T,HOU T,et al. A hybrid short-term load forecasting model based on improved fuzzy C-means clustering,random forest and deep neural networks[J].IEEE Access,2021,9:59754-59765.
[13] FONTANINI A D,ABREU J. A data-driven BIRCH clustering method for extracting typical load profiles for big data[C]∥2018IEEE Power&Energy Society General Meeting(PESGM).Portland,OR,USA.IEEE,2018:1-5.
[14] CHEN Y W,ZHOU L D,PEI S W,et al.KNN-BLOCK DBSCAN:fast clustering for large-scale data[J]. IEEE Transactions on Systems,Man,and Cybernetics:Systems,2021,51(6):3939-3953.
[15]瞿原，邓维斌，胡峰，等.基于Spark的点排序识别聚类结构算法[J].计算机科学，2018,45(1):97-102.QU Yuan,DENG Weibin,HU Feng,et al. Algorithm for ordering points to identify clustering structure based on spark[J].Computer Science,2018,45(1):97-102.
[16]张红斌，贺仁睦，刘应梅.基于KOHONEN神经网络的电力系统负荷动特性聚类与综合[J].中国电机工程学报，2003,23(5):1-5.ZHANG Hongbin,HE Renmu,LIU Yingmei. The characteristics clustering and synthesis of electric dynamic loads based on kohonen neural network[J]. Proceedings of the CSEE,2003,23(5):1-5.
[17]黄乾，马开刚，韦善阳，等.基于强化自组织映射和径向基神经网络的短期负荷预测[J].全球能源互联网，2019,2(1):70-77.HUANG Qian,MA Kaigang,WEI Shanyang,et al. A short-term load forecasting method based on reinforcement self-organizing map and radial basis function neural network[J].Journal of Global Energy Interconnection,2019,2(1):70-77.
[18]王继东，顾志成，葛磊蛟，等.改进萤火虫算法与K-means算法结合的配电网负荷聚类特性分析[J].天津大学学报（自然科学与工程技术版），2023,56(2):137-147.WANG Jidong,GU Zhicheng,GE Leijiao,et al. Load clustering characteristic analysis of the distribution network based on the combined improved firefly algorithm and K-means algorithm[J].Journal of Tianjin University(Science and Technology),2023,56(2):137-147.
[19]谭若洋，王治宇.基于K-means聚类的改进粒子群优化算法研究[J].统计与咨询，2020(3):24-27.TAN Ruoyang,WANG Zhiyu. Research on improved particle swarm optimization algorithm based on K-means clustering[J].Statistics and Consultation,2020(3):24-27.
[20]孙景钌，胡长洪，项烨鋆，等.基于多核模糊C均值聚类的配电网短期负荷预测[J].浙江电力，2022,41(3):65-71.SUN Jingliao,HU Changhong,XIANG Yeyun,et al. Short-term load forecasting of distribution networks based on multiple kernel fuzzy C-means clustering[J]. Zhejiang Electric Power,2022,41(3):65-71.
[21]梁京章，黄星舒，吴丽娟，等.基于KPCA和改进K-means的电力负荷曲线聚类方法[J].华南理工大学学报（自然科学版），2020,48(6):143-150.LIANG Jingzhang,HUANG Xingshu,WU Lijuan,et al.Clustering method of power load profiles based on KPCA and improved Kmeans[J]. Journal of South China University of Technology(Natural Science Edition),2020,48(6):143-150.
[22]白雅玲，周亚同，刘君.基于深度卷积嵌入聚类的日负荷曲线聚类分析[J].电网技术，2022,46(6):2104-2113.BAI Yaling,ZHOU Yatong,LIU Jun. Clustering analysis of daily load curve based on deep convolution embedding clustering[J].Power System Technology,2022,46(6):2104-2113.
[23]王连栋.城市网格化智能配电网的规划分析[J].电子技术，2023,52(8):56-57.WANG Liandong. Analysis of planning for urban grid intelligent distribution network[J]. Electronic Technology,2023,52(8):56-57.
[24]张吉祥.基于能源互联网的城市配电网规划优化方法研究[D].北京：华北电力大学，2017.
[25]殷丽凤，栗庆杰.启发式k-means聚类算法的改进研究[J].大连交通大学学报，2024,45(2):115-119.YIN Lifeng,LI Qingjie.Study on improvement of heuristic k-means clustering algorithm[J]. Journal of Dalian Jiaotong University,2024,45(2):115-119.
[26] GOODFELLOW I,POUGET-ABADIE J,MIRZA M,et al.Generative adversarial networks[J].Communications of the ACM,2020,63(11):139-144.
[27]谢天圻，吴媛媛，敬超，等．GAN模型生成图像检测方法综述[J/OL]．计算机工程与应用，1-15[2024-09-20].XIE Tianqi,WU Yuanyuan,JING Chao,et al. A review of GAN model generation image detection methods[J/OL]. Computer Engineering and Applications,1-15[2024-09-20].
[28]邵振国，张承圣，陈飞雄，等.生成对抗网络及其在电力系统中的应用综述[J].中国电机工程学报，2023,43(3):987-1003.SHAO Zhenguo,ZHANG Chengsheng,CHEN Feixiong,et al. A review on generative adversarial networks for power system applications[J].Proceedings of the CSEE,2023,43(3):987-1003.
[29]时纯，刘君，梁卓航，等.基于GAN和多通道CNN的电力系统暂态稳定评估[J].电网技术，2022,46(8):3191-3202.SHI Chun,LIU Jun,LIANG Zhuohang,et al. Transient stability assessment of power system based on GAN and multi-channel CNN[J].Power System Technology,2022,46(8):3191-3202.
[30] MUKHERJEE S,ASNANI H,LIN E,et al. ClusterGAN:latent space clustering in generative adversarial networks[J]. ArXiv ePrints,2018:arXiv:1809.03627.
[31]王磊，张野，张雅婷，等.考虑量测数据缺失的电力系统暂态稳定评估[J].电气自动化，2022,44(6):22-24.WANG Lei,ZHANG Ye,ZHANG Yating,et al.Transient stability assessment of PowerSystem considering missing measurement data[J].Electrical Automation,2022,44(6):22-24.
基本信息:
DOI：10.20097/j.cnki.issn1007-9904.240328
中图分类号:TM714
引用信息:
[1]徐春雷,焦昊,马洲俊,等.基于改进K-means和Cluster-GAN的配电网网格划分与负荷聚类[J].山东电力技术,2025,52(11):52-66.DOI:10.20097/j.cnki.issn1007-9904.240328.
基金信息:
国网江苏省电力有限公司科技项目“数字电网精准映射与计算推演技术研究”（J2023121）~~