基于Transformer深度学习网络的主动配电网多元源荷灾损辨识方法

doi:10.12204/j.issn.1000-7229.2023.03.006

电力建设 ›› 2023, Vol. 44 ›› Issue (3): 56-65.doi: 10.12204/j.issn.1000-7229.2023.03.006

• 新型电力系统下配电网规划与运行优化关键技术研究及应用·栏目主持王守相教授、赵倩宇博士· • 上一篇下一篇

基于Transformer深度学习网络的主动配电网多元源荷灾损辨识方法

纪鹏志¹, 李光肖¹, 王琳¹, 刘思贤¹, 刘宗杰¹, 马梓耀², 王照琪²(), 唐巍²()

1.国网山东省电力公司济宁供电公司,山东省济宁市 272100
2.中国农业大学信息与电气工程学院,北京市 100083

收稿日期:2022-03-10 出版日期:2023-03-01 发布日期:2023-03-02
通讯作者: 王照琪(1997),女,硕士,博士研究生,研究方向为配电网规划仿真、弹性配电网、电动汽车,E-mail:Seven_wzq@cau.edu.cn
作者简介:纪鹏志(1978),男,硕士,高级工程师,主要研究方向为配电网规划与运行分析
李光肖(1981),男,学士,高级工程师,主要研究方向为配电网规划与运行分析
王琳(1987),男,学士,高级工程师,主要研究方向为配电网规划与运行分析
刘思贤(1989),男,硕士,高级工程师,主要研究方向为主动配电网规划
刘宗杰(1983),男,学士,高级工程师,主要研究方向为主动配电网规划
马梓耀(2000),男,硕士研究生,主要研究方向为弹性配电网规划
唐巍(1971),女,博士,教授,博士研究生导师,研究方向为配电网规划与评估、配电网经济安全运行、分布式发电与微电网技术,E-mail:wei_tang@cau.edu.cn
基金资助:
国网山东省电力公司科技项目“面向多元源荷接入的主动配电网多目标协调规划技术研究”(5206002000R2)

Transformer-Based Evaluation Method of Power Outage in Active Distribution Networks with Multiple Sources and Loads

JI Pengzhi¹, LI Guangxiao¹, WANG Lin¹, LIU Sixian¹, LIU Zongjie¹, MA Ziyao², WANG Zhaoqi²(), TANG Wei²()

1. State Grid Jining Electric Power Supply Company, Jining 272100, Shandong Province, China
2. College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China

Received:2022-03-10 Online:2023-03-01 Published:2023-03-02
Supported by:
State Grid Shandong Electric Power Company Research Program(5206002000R2)

摘要/Abstract

摘要：

近年来不断增多的强台风天气给沿海及部分内陆地区配电网带来了愈发严重的损失,造成大规模重要负荷长时间失电,提高含多元源荷的主动配电网恢复能力成为亟待解决的问题。针对现有配电网负荷损失评估方法在强台风弱通信条件下无法准确获取节点信息而造成灾损分析精度不高的问题,提出一种基于Transformer深度学习网络的主动配电网多元源荷灾损辨识方法,充分利用深度学习网络去模型化的特点并发挥其在灾损预测精度方面的优势。考虑地面粗糙程度和高度,结合弱通信条件下的台风灾害气象数据,构建主动配电网所处地理环境的风速、降雨强度等气象信息修正模型;在此基础上考虑强台风致灾机理和主动配电网拓扑结构,利用Transformer深度学习方法构建配电网灾损辨识模型,实现强台风弱通信条件下的主动配电网多元源荷灾损辨识精度提升。通过对改进的IEEE 33节点主动配电网算例进行仿真测试,对损失负荷、损坏节点数等特征量进行计算,验证了所提主动配电网多元源荷灾损辨识方法能够满足台风多发配电网灾损评估精度。

关键词: 主动配电网, 多元源荷, 台风, 灾损辨识, 深度学习

Abstract:

In recent years, the increasing strong typhoon weather has brought more and more serious losses to the distribution network in coastal and some inland areas, resulting in large-scale power loss of important load for a long time. Improving the recovery capacity of active distribution networks containing multiple sources and loads has become an urgent problem to be solved. To deal with the problem that the existing power outage evaluation methods of distribution networks failed to obtain the information of the nodes, this paper proposes a novel evaluation method of power outage in active distribution networks containing multiple sources and loads on the basis of deep learning method, such as the Transformer model. Considering the ground roughness and height, meteorological information correction model of the geographical environment of the active distribution network is constructed combined with the typhoon disaster meteorological data under the condition of weak communication. On this basis, considering the disaster mechanism of strong typhoons and the topology of active distribution networks, the Transformer model is used to construct the power outage model of active distribution networks to improve the accuracy of power outage evaluation in active distribution networks containing multiple sources and loads. Through the simulation tests of the improved IEEE 33-node active distribution network, it is verified that the proposed power outage evaluation method for active distribution networks can meet the accuracy requirement of power outage evaluation in typhoon-prone distribution networks.

Key words: active distribution network, multiple sources and loads, typhoon, power outage evaluation, deep learning

中图分类号:

TM73

纪鹏志, 李光肖, 王琳, 刘思贤, 刘宗杰, 马梓耀, 王照琪, 唐巍. 基于Transformer深度学习网络的主动配电网多元源荷灾损辨识方法[J]. 电力建设, 2023, 44(3): 56-65.

JI Pengzhi, LI Guangxiao, WANG Lin, LIU Sixian, LIU Zongjie, MA Ziyao, WANG Zhaoqi, TANG Wei. Transformer-Based Evaluation Method of Power Outage in Active Distribution Networks with Multiple Sources and Loads[J]. ELECTRIC POWER CONSTRUCTION, 2023, 44(3): 56-65.

图/表 11

图1 基于Transformer的节点失电预测方法

Fig.1 Transformer-based prediction method of node power loss

表1 地形特征参考值

Table 1 Reference values of terrain features

表2 不同地形粗糙度对应表

Table 2 Corresponding table of different terrain roughness

图2 基于Transformer预测模型的结构

Fig.2 Structure of Transformer-based prediction model

图3 改进的IEEE 33节点配电网拓扑

Fig.3 Improved IEEE 33-node distribution network topology

图4 考虑不同气象因素的节点预测误差

Fig.4 Node prediction error considering different meteorological factors

图5 机理修正前的拓扑故障识别结果

Fig.5 Topological fault identification result before mechanism correction

图6 机理修正后的拓扑故障识别结果

Fig.6 Topological fault identification result after mechanism modification

表3 自适应算例结果对比

Table 3 Comparison of adaptive example results

图7 Transformer与LSTM预测误差对比

Fig.7 Comparison of prediction errors between Transformer-based and LSTM methods

图8 济宁市10 kV嘉因线模拟算例故障图

Fig.8 Fault diagram of Jining 10 kV Jiayin line simulation example

参考文献 30

[1]	侯慧, 俞菊芳, 耿浩, 等. 台风灾害下配网用户停电数量预测最优数据驱动模型选择[J]. 电力系统保护与控制, 2021, 49(13):114-120.
	HOU Hui, YU Jufang, GENG Hao, et al. Selection of optimal data-driven model for forecasting outage number of distribution network users under typhoon disaste[J]. Power System Protection and Control, 2021, 49(13):114-120.
[2]	侯慧, 陈希, 李敏, 等. 一种Stacking集成结构的台风灾害下停电空间预测方法[J]. 电力系统保护与控制, 2022, 50(3):76-84.
	HOU Hui, CHEN Xi, LI Min, et al. A space prediction method for power outage in a typhoon disaster based on a Stacking integrated structure[J]. Power System Protection and Control, 2022, 50(3):76-84.
[3]	姜萌萌, 佟永吉, 陈明丰, 等. 基于灾害条件的主动配电网快速重构策略研究[J]. 智慧电力, 2021, 49(5):85-92.
	JIANG Mengmeng, TONG Yongji, CHEN Mingfeng, et al. Rapid reconfiguration strategy of active distribution network under disaster conditions[J]. Smart Power, 2021, 49(5):85-92..
[4]	李振坤, 王法顺, 郭维一, 等. 极端天气下智能配电网的弹性评估[J]. 电力系统自动化, 2020, 44(9): 60-68.
	LI Zhenkun, WANG Fashun, GUO Weiyi, et al. Resilience evaluation of smart distribution network in extreme weather[J]. Automation of Electric Power Systems, 2020, 44(9): 60-68.
[5]	LIU Y, SINGH C. A methodology for evaluation of hurricane impact on composite power system reliability[J]. IEEE Transactions on Power Systems, 2011, 26(1): 145-152. doi: 10.1109/TPWRS.2010.2050219 URL
[6]	王红君, 陈智晴, 赵辉, 等. 考虑风光荷不确定性的配电网故障恢复策略[J]. 电网技术, 2022, 46(11): 4356-4363.
	WANG Hongjun, CHEN Zhiqing, ZHAO Hui, et al. Reconstruction strategies for fault recovery of active distribution network with distributed generation uncertainties[J]. Power System Technology, 2022, 46(11): 4356-4363.
[7]	吴勇军, 薛禹胜, 谢云云, 等. 台风及暴雨对电网故障率的时空影响[J]. 电力系统自动化, 2016, 40(2): 20-29, 83.
	WU Yongjun, XUE Yusheng, XIE Yunyun, et al. Space-time impact of typhoon and rainstorm on power grid fault probability[J]. Automation of Electric Power Systems, 2016, 40(2): 20-29, 83.
[8]	李锐, 陈颖, 梅生伟, 等. 基于停电风险评估的城市配电网应急预警方法[J]. 电力系统自动化, 2010, 34(16): 19-23.
	LI Rui, CHEN Ying, MEI Shengwei, et al. An early warning method for emergency response based on power failure risk analysis of distribution systems[J]. Automation of Electric Power Systems, 2010, 34(16): 19-23.
[9]	BATTISTA R C, RODRIGUES R S, PFEIL M S. Dynamic behavior and stability of transmission line towers under wind forces[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2003, 91(8): 1051-1067. doi: 10.1016/S0167-6105(03)00052-7 URL
[10]	孙春红, 赵艳雷. 基于设备脆弱性的电网气象灾害故障预测[J]. 山东电力技术, 2020, 47(2): 9-12.
	SUN Chunhong, ZHAO Yanlei. Meteorological disaster fault prediction for power grid based on equipment vulnerability[J]. Shandong Electric Power, 2020, 47(2): 9-12.
[11]	陈彬, 于继来. 强台风环境下配电网断杆概率的网格化评估[J]. 电气应用, 2018, 37(16): 42-47.
	CHEN Bin, YU Jilai. Grid-based evaluation of the probability of broken poles in distribution network under strong typhoon environment[J]. Electrotechnical Application, 2018, 37(16): 42-47.
[12]	王中山, 袁金锦. 面向灾害风险评估的台风数值模拟及可视化[J]. 测绘通报, 2015(4): 108-110.
	WANG Zhongshan, YUAN Jinjin. Typhoon numerical simulation and visualization for disaster risk assessment[J]. Bulletin of Surveying and Mapping, 2015(4): 108-110.
[13]	王永明, 殷自力, 李琳, 等. 台风灾害场景下考虑运行状态的配电网风险评估方法[J]. 电力系统及其自动化学报, 2018, 30(12): 60-65.
	WANG Yongming, YIN Zili, LI Lin, et al. Risk assessment method for distribution network considering operation state in the scene of typhoon disaster[J]. Proceedings of the CSU-EPSA, 2018, 30(12): 60-65.
[14]	SIJIVAC D, NIKOLOVSKI S, KOVAC Z. Distribution network restoration using sequential Monte Carlo approach[C]// 2006 International Conference on Probabilistic Methods Applied to Power Systems. IEEE, 2007: 1-6.
[15]	LIN D, LIU Q J, LI Z H, et al. Elaborate reliability evaluation of cyber physical distribution systems considering fault location, isolation and supply restoration process[J]. IEEE Access, 2020, 8: 128574-128590. doi: 10.1109/Access.6287639 URL
[16]	ZHANG Z W, HOU H, LI M, et al. Weighted failure probability calculation of overhead distribution line in random wind field[C]// 2020 IEEE Symposium Series on Computational Intelligence (SSCI). IEEE, 2021: 968-975.
[17]	LIU P, WANG M, SUN C, et al. Research of the disaster situation impact on low voltage urban distribution grid’s power supply reliability[C]// 2018 China International Conference on Electricity Distribution (CICED). IEEE, 2018: 1031-1035.
[18]	KALYSH I, KENZHINA M, KAIYRBEKOV N, et al. Machine learning-based service restoration scheme for smart distribution systems with DGs and high priority loads[C]// 2019 International Conference on Smart Energy Systems and Technologies (SEST). IEEE, 2019: 1-6.
[19]	FU J F, NÚÑEZ A, DE SCHUTTER B. Real-time UAV routing strategy for monitoring and inspection for postdisaster restoration of distribution networks[J]. IEEE Transactions on Industrial Informatics, 2022, 18(4): 2582-2592. doi: 10.1109/TII.2021.3098506 URL
[20]	熊小伏, 王尉军, 于洋, 等. 多气象因素组合的输电线路风险分析[J]. 电力系统及其自动化学报, 2011, 23(6): 11-15.
	XIONG Xiaofu, WANG Weijun, YU Yang, et al. Risk analysis method for transmission line combining of various meteorological factors[J]. Proceedings of the CSU-EPSA, 2011, 23(6): 11-15.
[21]	朱燕芳, 闫磊, 常康, 等. 基于深度卷积神经网络的电力系统故障预测[J/OL]. 电源学报. [2022-03-10]. http://kns.cnki.net/kcms/detail/12.1420.TM.20220328.1722.006.html.
	ZHU Yanfang, YAN Lei, CHANG Kang, et al. Power system fault prediction based on deep convolutional neural network[J/OL]. Journal of Power Supply.[2022-03-10]. http://kns.cnki.net/kcms/detail/12.1420.TM.20220328.1722.006.html.
[22]	丁岩, 郭和强, 王利利, 等. 基于隐马尔可夫模型的电力系统连锁故障预测[J]. 电网与清洁能源, 2017, 33(9): 93-98.
	DING Yan, GUO Heqiang, WANG Lili, et al. Power system cascading failure prediction based on hidden Markov model[J]. Power System and Clean Energy, 2017, 33(9): 93-98.
[23]	严文君, 刘俊勇. 基于深度学习的多微电网市场主从博弈运行优化策略[J]. 电力系统及其自动化学报, 2022, 34(7): 120-128.
	YAN Wenjun, LIU Junyong. Optimization strategy for master-slave game operation in multi-microgrid market based on deep learning[J]. Proceedings of the CSU-EPSA, 2022, 34(7): 120-128.
[24]	俞小勇, 秦丽文, 桂海涛, 等. 新一代人工智能在配电网智能感知与故障诊断中的应用[J]. 南方电网技术, 2022, 16(5): 34-43.
	YU Xiaoyong, QIN Liwen, GUI Haitao, et al. Application of new generation artificial intelligence technology in intelligent sensing and fault detection of power distribution network[J]. Southern Power System Technology, 2022, 16(5): 34-43.
[25]	袁春红, 杨振斌, 薛桁, 等. 复杂地形风速数值模拟[J]. 太阳能学报, 2002, 23(3): 374-377.
	YUAN Chunhong, YANG Zhenbin, XUE Heng, et al. The numerical simulation on wind speed over complex topogtraphy[J]. Acta Energiae Solaris Sinica, 2002, 23(3): 374-377.
[26]	李正昊, 楼文娟, 章李刚, 等. 地貌因素对垭口内风速影响的数值模拟[J]. 浙江大学学报(工学版), 2016, 50(5): 848-855.
	LI Zhenghao, LOU Wenjuan, ZHANG Ligang, et al. Numerical simulation of effects of topographic factors on wind speed in col[J]. Journal of Zhejiang University (Engineering Science), 2016, 50(5): 848-855.
[27]	LI S Y, JIN X Y, XUAN Y, et al. Enhancing the locality and breaking the memory bottleneck of transformer on time series forecasting[C/OL]// 33rd Conference on Neural Information Processing Systems (NeurIPS 2019). [2022-03-10]. https://proceedings.neurips.cc/paper/2019/file/6775a0635c302542da2c32aa19d86be0-Paper.pdf.
[28]	WU N, GREEN B, BEN X, et al. Deep transformer models for time series forecasting: the influenza prevalence case[EB/OL]. [2022-03-10].2020: arXiv: 2001.08317. https://arxiv.org/abs/2001.08317.
[29]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all You need[EB/OL]. [2022-03-10]. 2017: arXiv: 1706.03762. https://arxiv.org/abs/1706.03762.
[30]	LI Y D, ZHANG L, LV Z, et al. Detecting anomalies in intelligent vehicle charging and station power supply systems with multi-head attention models[J]. IEEE Transactions on Intelligent Transportation Systems, 2021, 22(1): 555-564. doi: 10.1109/TITS.6979 URL

基于Transformer深度学习网络的主动配电网多元源荷灾损辨识方法

Transformer-Based Evaluation Method of Power Outage in Active Distribution Networks with Multiple Sources and Loads

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 30

相关文章 15

编辑推荐

Metrics

本文评价

[1]	谭媛, 张文海, 王杨. 基于多源信息融合的配网故障外部原因识别[J]. 电力建设, 2023, 44(3): 77-84.
[2]	李军徽, 马得轩, 朱星旭, 李翠萍, 侯涛. 基于ADMM算法的主动配电网分层优化经济调度[J]. 电力建设, 2022, 43(8): 76-86.
[3]	肖发龙, 吴岳忠, 沈雪豪, 何震凯, 秦烨. 基于深度学习和知识图谱的变电站设备故障智能诊断[J]. 电力建设, 2022, 43(3): 66-74.
[4]	孙黎霞, 彭嘉杰, 田屹昀, 陈欣凌, 袁欢. 基于混合门控循环单元子层的多任务暂态稳定评估[J]. 电力建设, 2022, 43(2): 63-69.
[5]	席向东, 侯香臣, 萨仁高娃, 朱星旭, 孙哲彬, 梁涛, 刘鹏宣, 王晓櫆, 李翠萍, 李军徽. 计及电压预测的主动配电网分布式电源在线优化决策方法[J]. 电力建设, 2022, 43(11): 24-32.
[6]	孟令卓超, 杨锡运, 赵泽宇. 考虑光-荷不确定性和旋转备用约束的主动配电网经济优化调度策略[J]. 电力建设, 2022, 43(11): 63-72.
[7]	邓靖微, 李华强, 温丰瑞, 王俊翔, 谢康胜, 金智博. 计及虚拟电厂市场交易的主动配电网两阶段优化调度[J]. 电力建设, 2021, 42(9): 22-31.
[8]	孙建梅, 胡嘉栋, 蔚芳. 考虑利益主体合作的分布式电源运营商电源优化配置[J]. 电力建设, 2021, 42(6): 127-134.
[9]	何惠清, 钱赟, 何炳源, 何乐彰, 王彬, 周哲民. 基于机会约束的主动配电网动态经济调度方法[J]. 电力建设, 2021, 42(4): 1-8.
[10]	朱庆, 郑红娟, 唐子逸, 韦思雅, 邹子骁, 吴熙. 基于生成对抗网络的综合能源负荷场景生成方法[J]. 电力建设, 2021, 42(12): 1-8.
[11]	唐子卓, 刘洋, 许立雄, 郭久亿. 基于负荷数据频域特征和LSTM网络的类别不平衡负荷典型用电模式提取方法[J]. 电力建设, 2020, 41(8): 17-24.
[12]	沈青文，程若发，付鑫. 一种计及脆弱性指标的分布式电源优化配置方法[J]. 电力建设, 2020, 41(3): 54-61.
[13]	戴松灵，王晞，刘方，雷云凯，欧阳雪彤，温丰瑞，李华强. 考虑综合承载力的主动配电网优化调度与运行[J]. 电力建设, 2020, 41(2): 67-75.
[14]	陈倩，齐林海，王红. 基于LSTM网络的谐波多标签分类[J]. 电力建设, 2019, 40(7): 101-106.
[15]	肖泽青，华昊辰，曹军威. 人工智能在能源互联网中的应用综述[J]. 电力建设, 2019, 40(5): 63-70.