| 591 | 0 | 835 |
| 下载次数 | 被引频次 | 阅读次数 |
随着分布式能源的广泛接入，配电网运行状态日益复杂，对状态估计精度提出更高要求。然而，传统状态估计方法往往依赖于预设的拓扑参数信息，难以应对网络结构变化和参数不确定性，导致估计精度下降。为此，提出一种基于数据驱动拓扑参数辨识的配电网状态估计方法。首先，通过区域划分减少量测数据需求，并结合LinDistFlow潮流模型构建拓扑标识矩阵，利用矩阵合同变换实现拓扑和线路参数的联合辨识；其次，引入Kmeans++聚类算法提升模型鲁棒性，并考虑网络损耗优化线路参数估计；然后，基于精确的拓扑和参数信息，采用贝叶斯优化的卷积神经网络进行状态估计；最后，通过在改进后的IEEE 33和IEEE 123节点网络上的仿真实验，验证了所提方法在不同噪声和数据规模下均表现出较高的准确性和鲁棒性。
Abstract：With the widespread integration of distributed energy,the operating states of distribution networks are becoming increasingly complex,placing higher demands on the accuracy of state estimation. However,the traditional state estimation methods often rely on predefined topology and parameter information,making it difficult to adapt to changes in network structure and parameter uncertainty and leading to reduced estimation accuracy. To address this issue,a state estimation method for distribution network based on data-driven topology and parameter identification is proposed in this paper. First,measurement data requirements are reduced through regional partitioning,a topology label matrix is constructed by combining the LinDistFlow power flow model,and matrix contract transformation is used to achieve a joint identification of topology and line parameters. Second,the K-means++ clustering algorithm is introduced to enhance the model robustness,and network losses are considered to optimize the line parameter estimation. Third,based on the precise topology and parameter information,a Bayesian-optimized convolutional neural network is employed for state estimation. Finally,the proposed method's accuracy and robustness are validated through simulation experiments,which are conducted on modified IEEE 33-node and 123-node networks under conditions of different noises and different data scales.
[1]黄悦华，涂金童，陈庆，等（Huang Yuehua,Tu Jintong,Chen Qing,et al）.基于场景模糊聚类和多主体演化博弈的主动配电网协同规划（Active distribution network collaborative planning based on scenario fuzzy clustering and multi-agent evolutionary game）[J].电力系统及其自动化学报（Proceedings of the CSU-EPSA）,2025,37(5):52-66.
[2]龙寰，石子晴，赵景涛，等（Long Huan,Shi Ziqing,Zhao Jingtao,et al）.基于多标签分类与卷积神经网络的配电网拓扑辨识（Topology identification of distribution network based on multi-label classification and CNN）[J].高电压技术（High Voltage Engineering）,2024,50(10):4520-4529.
[3]杨秀，蒋家富，刘方，等（Yang Xiu,Jiang Jiafu,Liu Fang,et al）.基于注意力机制和卷积神经网络的配电网拓扑辨识（Distribution network topology identification based on attention mechanism and convolutional neural network）[J].电网技术（Power System Technology）,2022,46(5):1672-1682.
[4]张汪洋，樊艳芳，侯俊杰，等（Zhang Wangyang,Fan Yanfang,Hou Junjie,et al）.基于集成深度神经网络的配电网分布式状态估计方法（Distribution network distributed state estimation method based on an integrated deep neural network）[J].电力系统保护与控制（Power System Protection and Control）,2024,52(3):128-140.
[5]田钧祥，陈铁，陈彬（Tian Junxiang,Chen Tie,Chen Bin）.基于改进自适应UKF算法的中压配电网鲁棒动态状态估计方法（Robust dynamic state estimation method for medium voltage distribution networks based on improved adaptive UKF algorithm）[J].中国电力（Electric Power）,2023,56(11):128-133.
[6]龚逊东，费有蝶，凌佳凯，等（Gong Xundong,Fei Youdie,Ling Jiakai,et al）.基于深度神经网络伪量测建模的交直流混合配电网交替迭代法状态估计（Alternating-iteration state estimation of AC/DC hybrid distribution system based on pseudo-measurement modeling using deep neural network）[J].电力建设（Electric Power Construction）,2022,43(10):111-120.
[7]吉兴全，姜海洋，张玉敏，等（Ji Xingquan,Jiang Haiyang,Zhang Yumin,et al）.考虑网络重构的三相不平衡配电网μPMU组合优化配置（Combined optimal configuration ofμPMU for three-phase unbalanced distribution network considering network reconstruction）[J].电力系统自动化（Automation of Electric Power Systems）,2025,49(3):145-155.
[8]王献志，李铁成，郭少飞，等（Wang Xianzhi,Li Tiecheng,Guo Shaofei,et al）.基于μPMU电压相位数据的配电网拓扑辨识（Identification of distribution network topology based onμPMU voltage phase data）[J].电网与清洁能源（Power System and Clean Energy）,2022,38(6):54-63.
[9] Srinivas V L,Wu Jianzhong. Topology and parameter identification of distribution network using smart meter andμPMU measurements[J]. IEEE Transactions on Instrumentation and Measurement,2022,71:9004114.
[10] Shah P,Zhao Xiaowei. Network identification usingμPMU and smart meter measurements[J]. IEEE Transactions on Industrial Informatics,2022,18(11):7572-7586.
[11]王贺飞，李洪涛，余昆，等（Wang Hefei,Li Hongtao,Yu Kun,et al）.基于DSCADA和μPMU数据融合的配电网运行拓扑辨识（Identification of distribution network operation topology based on DSCADA andμPMU telemetry data fusion）[J].电力建设（Electric Power Construction）,2019,40(6):123-131.
[12] Zhang Jiawei,Wang Yi,Weng Yang,et al. Topology identification and line parameter estimation for non-PMU distribution network:A numerical method[J]. IEEE Transactions on Smart Grid,2020,11(5):4440-4453.
[13]宁家鑫，刘羽霄，章家维，等（Ning Jiaxin,Liu Yuxiao,Zhang Jiawei,et al）.数据驱动的三相配电网络拓扑与线路参数辨识（Data-driven topology and line parameter identification of three-phase distribution grid）[J].中国电机工程学报（Proceedings of the CSEE）,2021,41(8):2615-2627.
[14]杨冬锋，付强，刘晓军，等（Yang Dongfeng,Fu Qiang,Liu Xiaojun,et al）.配电网动态拓扑与线路参数联合在线辨识方法（Joint online identification method for dynamic topology and line parameters of distribution network）[J].电力系统自动化（Automation of Electric Power Systems）,2022,46(2):101-108.
[15]马尚，卫志农，黄蔓云，等（Ma Shang,Wei Zhinong,Huang Manyun,et al）.基于智能电表数据的低压配电网拓扑与线路参数联合辨识（Joint identification of topology and line parameters for low-voltage distribution network based on smart meter data）[J].电力系统自动化（Automation of Electric Power Systems）,2024,48(2):60-70.
[16]朱鹏程，柳劲松，范士雄，等（Zhu Pengcheng,Liu Jinsong,Fan Shixiong,et al）.考虑混合量测的配电网二次约束二次估计方法（A quadratic constraint quadratic estimation method based on hybrid measurements for distribution networks）[J].电网技术（Power System Technology）,2019,43(3):841-847.
[17]高岩，吴汉斌，张纪欣，等（Gao Yan,Wu Hanbin,Zhang Jixin,et al）.基于组合深度学习的光伏功率日前概率预测模型（Day-ahead probabilistic prediction model for photovoltaic power based on combined deep learning）[J].中国电力（Electric Power）,2024,57(4):100-110.
[18]李鹏，祖文静，刘一欣，等（Li Peng,Zu Wenjing,Liu Yixin,et al）.基于不完全量测数据的配电网状态估计方法（State estimation method for distribution network based on incomplete measurement data）[J].中国电力（Electric Power）,2025,58(5):1-10.
[19]夏添梁，张玉敏，杨明，等（Xia Tianliang,Zhang Yumin,Yang Ming,et al）.联合长短期记忆神经网络和粒子滤波的配电网预测辅助鲁棒状态估计方法（Robust forecasting-aided state estimation method of distribution network based on long-short term memory neural network and particle filter）[J].高电压技术（High Voltage Engineering）,2022,48(4):1343-1355.
[20]兰浦哲，韩冬，徐潇源，等（Lan Puzhe,Han Dong,Xu Xiaoyuan,et al）.基于长短期记忆的电-气耦合综合能源系统贝叶斯状态估计（Bayesian state estimation for electricity-gas coupled integrated energy system based on long short-term memory）[J].电力系统自动化（Automation of Electric Power Systems）,2021,45(20):18-28.
[21] Wang Xiaoxue,Zhao Yikang,Zhou Yue. A data-driven topology and parameter joint estimation method in nonPMU distribution networks[J]. IEEE Transactions on Power Systems,2024,39(1):1681-1692.
[22]张华幸（Zhang Huaxing）.基于高斯混合模型的代价敏感序贯三支人脸识别研究（Gaussian Mixture Model Based Cost Sensitive Sequential Three-Way Decision for Face Recognition）[D].南京：南京大学（Nanjing:Nanjing University）,2021.
[23] Koyama A,Miyauchi S,Morooka K,et al. Analysis of TEM images of metallic nanoparticles using convolutional neural networks and transfer learning[J]. Journal of Magnetism and Magnetic Materials,2021,538:168225.
基本信息:
DOI：10.19635/j.cnki.csu-epsa.001685
中图分类号:TM73
引用信息:
[1]梁俊宇,杨家全,赵翼康.基于数据驱动拓扑参数辨识的配电网状态估计方法[J].电力系统及其自动化学报,2025,37(11):96-104.DOI:10.19635/j.cnki.csu-epsa.001685.
基金信息:
中国南方电网有限责任公司创新项目（YNKJXM20222334,YNKJXM20222360）
2025-05-29
2025
2025-11-24
2025
2025-07-12
1
2025-07-18
2025-07-18
2025-07-18