
浏览全部资源
扫码关注微信
1.浙江工业大学信息工程学院,浙江省杭州市 310023
2.浙江大学电气工程学院,浙江省杭州市 310027
冯昌森(1990—),男,博士,副教授,主要研究方向:电力系统优化与控制、电力市场、机器学习等。E-mail:fcs@zjut.edu.cn.
钱燚飞(2000—),男,硕士,主要研究方向:电力系统优化控制、机器学习等。E-mail:493294587@qq.com
邵亮(2000—),男,博士研究生,主要研究方向:电力系统故障诊断与系统恢复。E-mail:shaol8849@gmail.com
张有兵(1971—),男,通信作者,教授,博士生导师,主要研究方向:智能电网、分布式发电及新能源优化控制、电动汽车入网、电力系统通信、电能质量监控。E-mail: youbingzhang@zjut.edu.cn
收稿日期:2024-08-19,
修回日期:2025-02-24,
网络出版日期:2025-06-10,
纸质出版日期:2025-08-25
移动端阅览
冯昌森,钱燚飞,邵亮等.考虑特征缺失的个性化居民短期负荷预测[J].电力系统自动化,2025,49(16):75-84.
FENG Changsen,QIAN Yifei,SHAO Liang,et al.Personalized Short-term Residential Load Forecasting Considering Missing Feature[J].Automation of Electric Power Systems,2025,49(16):75-84.
冯昌森,钱燚飞,邵亮等.考虑特征缺失的个性化居民短期负荷预测[J].电力系统自动化,2025,49(16):75-84. DOI: 10.7500/AEPS20240819005.
FENG Changsen,QIAN Yifei,SHAO Liang,et al.Personalized Short-term Residential Load Forecasting Considering Missing Feature[J].Automation of Electric Power Systems,2025,49(16):75-84. DOI: 10.7500/AEPS20240819005.
随着居民用电负荷在全社会总用电负荷中的占比逐渐提高,精准的居民负荷预测结果对提升电力系统运行的安全性和经济性,尤其是促进需求侧响应具有重要意义。然而,由于许多客观和人为因素,例如,量测配置不完整、量测设备故障和数据传输问题等,会导致负荷数据不全、丢失或畸变等问题。此外,居民负荷不确定性较大,要实现高精度负荷预测难度较大。在上述背景下,首先,提出一种个性化联邦学习框架,在不侵犯居民数据隐私的情况下,针对每个居民端定制高精度的个性化负荷预测模型。然后,采用图特征传播(GFP)方法重构缺失的负荷特征参数,可在相当程度上避免数据缺失引起的预测精度下降问题。最后,采用算例对所提方法进行测试。仿真结果表明,所提方法与现有代表性方法相比,其预测精度较高且处理数据缺失的性能也明显优于常用的插值方法。
As the proportion of residential electricity load in the total electricity load of the society gradually increases
accurate residential load forecasting results are of great significance for enhancing the safety and economy of the power system operation
especially for promoting demand-side response. However
due to many objective and human factors
such as incomplete measurement configuration
measurement equipment failures and data transmission problems
load data may be incomplete
lost or distorted. Moreover
residential load is highly uncertain
making it difficult to achieve high-precision load forecasting. Under this background
first
a personalized federated learning framework is proposed
which customizes high-precision personalized load forecasting models for each residential end without infringing on the privacy of residential data. Then
the graph feature propagation (GFP) method is used to reconstruct the missing load feature parameters
which can largely avoid the problem of forecasting accuracy decline caused by data missing. Finally
the proposed method is tested through cases. Simulation results show that
compared with the existing representative methods
the proposed method has higher forecasting accuracy and significantly better performance in handling data missing compared to common interpolation methods.
JAHAN I S , SNASEL V , MISAK S . Intelligent systems for power load forecasting: a study review [J]. Energies , 2020 , 13 ( 22 ): 6105 .
徐晶 , 迟福建 , 葛磊蛟 , 等 . 基于SARIMA-GRNN-SVM的短期商业电力负荷组合预测方法 [J]. 电力系统及其自动化学报 , 2020 , 32 ( 2 ): 85 - 91 .
XU Jing , CHI Fujian , GE Leijiao , et al . Short-term combined commercial load forecasting method based on SARIMA-GRNN-SVM [J]. Proceedings of the CSU-EPSA , 2020 , 32 ( 2 ): 85 - 91 .
谭风雷 , 张军 , 马宏忠 . 基于趋势变化分段的电力负荷组合预测方法 [J]. 华北电力大学学报(自然科学版) , 2020 , 47 ( 2 ): 17 - 24 .
TAN Fenglei , ZHANG Jun , MA Hongzhong . Combined forecasting method of power load based on trend change division [J]. Journal of North China Electric Power University (Natural Science Edition) , 2020 , 47 ( 2 ): 17 - 24 .
刘鑫蕊 , 常鹏 , 孙秋野 . 基于XGBoost和无迹卡尔曼滤波自适应混合预测的电网虚假数据注入攻击检测 [J]. 中国电机工程学报 , 2021 , 41 ( 16 ): 5462 - 5476 .
LIU Xinrui , CHANG Peng , SUN Qiuye . Grid false data injection attacks detection based on XGBoost and unscented Kalman filter adaptive hybrid prediction [J]. Proceedings of the CSEE , 2021 , 41 ( 16 ): 5462 - 5476 .
肖白 , 高文瑞 , 李道明 , 等 . 基于3σ-CEEMDAN-LSTM的空间负荷预测方法 [J]. 电力自动化设备 , 2023 , 43 ( 3 ): 159 - 165 .
XIAO Bai , GAO Wenrui , LI Daoming , et al . Spatial load forecasting method based on 3 <math id="M109"><mi mathvariant="normal">σ</mi></math> https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=87975728&type= 2.11666679 https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=87975750&type= 1.35466671 -CEEMDAN-LSTM [J]. Electric Power Automation Equipment , 2023 , 43 ( 3 ): 159 - 165 .
周念成 , 廖建权 , 王强钢 , 等 . 深度学习在智能电网中的应用现状分析与展望 [J]. 电力系统自动化 , 2019 , 43 ( 4 ): 180 - 191 .
ZHOU Niancheng , LIAO Jianquan , WANG Qianggang , et al . Analysis and prospect of deep learning application in smart grid [J]. Automation of Electric Power Systems , 2019 , 43 ( 4 ): 180 - 191 .
GONZALEZ R , AHMED S , ALAMANIOTIS M . Deep neural network based methodology for very-short-term residential load forecasting [C]// 2022 13th International Conference on Information , Intelligence, Systems & Applications, July 18-20, 2022 , Corfu, Greece : 1 - 6 .
杨强 . AI与数据隐私保护:联邦学习的破解之道 [J]. 信息安全研究 , 2019 , 5 ( 11 ): 961 - 965 .
YANG Qiang . AI and data privacy protection: the way to federated learning [J]. Journal of Information Security Research , 2019 , 5 ( 11 ): 961 - 965 .
周传鑫 , 孙奕 , 汪德刚 , 等 . 联邦学习研究综述 [J]. 网络与信息安全学报 , 2021 , 7 ( 5 ): 77 - 92 .
ZHOU Chuanxin , SUN Yi , WANG Degang , et al . Survey of federated learning research [J]. Chinese Journal of Network and Information Security , 2021 , 7 ( 5 ): 77 - 92 .
CHEN J X , GAO T L , SI R Q , et al . Residential short term load forecasting based on federated learning [C]// 2022 IEEE 2nd International Conference on Digital Twins and Parallel Intelligence , October 24-28 , 2022 , Boston, USA : 1 - 6 .
PARK K J , SON S Y . Residential load forecasting using modified federated learning algorithm [J]. IEEE Access , 2023 , 11 : 40675 - 40691 .
HE Y , LUO F J , RANZI G , et al . Short-term residential load forecasting based on federated learning and load clustering [C]// 2021 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids , October 25-28 , 2021 , Aachen, Germany : 77 - 82 .
HAN F J , PU T J , LI M Z , et al . Short-term forecasting of individual residential load based on deep learning and <math id="M110"><mi>K</mi></math> https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=87975751&type= 2.11666679 https://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=87975752&type= 2.37066650 -means clust ering [J]. CSEE Journal of Power and Energy Systems , 2021 , 7 ( 2 ): 261 - 269 .
LU N , LIU S , WEN Q S , et al . Federated domain separation for distributed forecasting of non-IID household loads [J]. IEEE Transactions on Smart Grid , 2024 , 15 ( 4 ): 4271 - 4283 .
WU D , LIN W X . Efficient residential electric load forecasting via transfer learning and graph neural networks [J]. IEEE Transactions on Smart Grid , 2023 , 14 ( 3 ): 2423 - 2431 .
HAMEED M Z , LAMICHHANE B . Transductive-transfer learning based deep neural networks for day-ahead PV power forecasting in smart grid application: a comparative analysis [C]// 2023 North American Power Symposium , October 15-17 , 2023 , Asheville, USA : 1 - 6 .
CAI Y H , LI M X , CHENG Y F . Small sample load forecasting of electric vehicle charging stations based on transfer learning [C]// 2024 5th International Seminar on Artificial Intelligence, Networking and Information Technology , March 29-31 , 2024 , Nanjing, China : 2041 - 2045 .
CHU W , SINDHWANI V , GHAHRAMANI Z , et al . Relational learning with Gaussian processes [M]. Cambridge, USA : The MIT Press , 2007 .
SHAMSIAN A , NAVON A , FETAYA E , et al . Personalized federated learning using hypernetworks [C]// 2021 38th International Conference on Machine Learning , July 18 - 24 , 2021 : 9489 - 9502 .
WANG Y , GAO N , HUG G . Personalized federated learning for individual consumer load forecasting [J]. CSEE Journal of Power and Energy Systems , 2023 , 9 ( 1 ): 326 - 330 .
CHEN Y , CHEN C Y , ZHANG X , et al . Privacy-preserving baseline load reconstruction for residential demand response considering distributed energy resources [J]. IEEE Transactions on Industrial Informatics , 2022 , 18 ( 5 ): 3541 - 3550 .
MA J , CHENG J C P , JIANG F F , et al . A bi-directional missing data imputation scheme based on LSTM and transfer learning for building energy data [J]. Energy and Buildings , 2020 , 216 : 109941 .
KIM T , KO W , KIM J . Analysis and impact evaluation of missing data imputation in day-ahead PV generation forecasting [J]. Applied Sciences , 2019 , 9 ( 1 ): 204 - 214 .
ZHAO J B , WANG Z Y , WANG J H . Robust time-varying load modeling for conservation voltage reduction assessment [J]. IEEE Transactions on Smart Grid , 2018 , 9 ( 4 ): 3304 - 3312 .
WIJAYA T K , VASIRANI M , ABERER K . When bias matters: an economic assessment of demand response baselines for residential customers [J]. IEEE Transactions on Smart Grid , 2014 , 5 ( 4 ): 1755 - 1763 .
ROSSI E , KENLAY H , GORINOVA M I , et al . On the unreasonable effectiveness of feature propagation in learning on graphs with missing node features [EB/OL].[ 2024-08-01 ]. https://arxiv.org/abs/2111.12128 https://arxiv.org/abs/2111.12128 .
王蓓蓓 , 朱竞 , 王嘉乐 , 等 . 电表数据隐私保护下的联邦学习行业电力负荷预测框架 [J]. 电力系统自动化 , 2023 , 47 ( 13 ): 86 - 93 .
WANG Beibei , ZHU Jing , WANG Jiale , et al . Federated-learning based industry load forecasting framework under privacy protection of meter data [J]. Automation of Electric Power Systems , 2023 , 47 ( 13 ): 86 - 93 .
李芬 , 李雨欣 , 王亚维 , 等 . 考虑建筑屋顶光伏热效应的净负荷短期预测 [J]. 电力系统自动化 , 2025 , 49 ( 9 ): 146 - 156 .
LI Fen , LI Yuxin , WANG Yawei , et al . Short-term forecasting of net load considering photovoltaic thermal effect on building roofs [J]. Automation of Electric Power Systems , 2025 , 49 ( 9 ): 146 - 156 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621