基于强化学习的电力通信网故障恢复方法

doi:10.11930/j.issn.1004-9649.201907078

中国电力 ›› 2020, Vol. 53 ›› Issue (6): 34-40.DOI: 10.11930/j.issn.1004-9649.201907078

• 人工智能在电力系统的应用 • 上一篇下一篇

基于强化学习的电力通信网故障恢复方法

贾惠彬, 盖永贺, 李保罡, 郑宏达

华北电力大学电气与电子工程学院，河北保定 071003

收稿日期:2019-07-09 修回日期:2019-08-26 发布日期:2020-06-05
作者简介:贾惠彬(1979-),男,博士,副教授,从事智能电网通信研究,E-mail:huibin.jia@foxmail.com;盖永贺(1995-),男,通信作者,硕士研究生,从事智能电网通信研究,E-mail:541250764@qq.com;李保罡(1980-),男,博士,副教授,从事无线通信、工业物联网、能源互联网、大数据分析研究,E-mail:baogangli@ncepu.edu.cn;郑宏达(1993-),男,硕士研究生,从事智能电网通信研究,E-mail:493533872@qq.com
基金资助:
国家自然科学基金资助项目(61971190)；中央高校基本科研业务费专项资金资助项目(2017MS113)

Power Communication Network Recovery from Large-Scale Failures Based on Reinforcement Learning

JIA Huibin, GAI Yonghe, LI Baogang, ZHENG Hongda

School of Electrical & Electronic Engineering, North China Electric Power University, Baoding 071003, China

Received:2019-07-09 Revised:2019-08-26 Published:2020-06-05
Supported by:
This work is supported by the National Natural Science Foundation of China(No.61971190), the Central University Basic Research Fees Special Fund Support Project (No.2017MS113)

摘要/Abstract

摘要： 大规模的自然灾害或恶意攻击会引发智能电网中电力通信网的严重故障，如果不能及时恢复，将给电力系统的安全稳定运行带来极大的风险。为解决电力通信网络大规模故障后的恢复问题，在通信链路恢复资源有限的约束下，建立以失效业务数量恢复最大化为目标的电力通信网大规模故障后链路恢复模型。针对该模型，提出一种基于强化学习的启发式算法，该算法利用链路恢复资源和故障链路在失效业务中的重要度，设置奖惩函数和选择规则，并累积奖励最大值，得到最优链路恢复组合。实验结果表明，提出的电力通信网故障恢复方法可以在恢复资源有限的约束下，恢复较多的失效业务。

关键词: 智能电网, 电力通信网, 大规模故障, 强化学习, 故障恢复

Abstract: Natural disasters or malicious attacks may cause large-scale failures of power communication networks in smart grid systems, which will impose big risks on the security and stability of the power system operation unless the communication network is recovered immediately. In order to solve the recovery problem of power communication network after large-scale failure, under the constraints of limited link recovery resources, a link recovery model for large-scale failures in power communication networks is established with the objective to maximize the recovery amount of failed services. Regarding this model, a heuristic algorithm based on reinforcement learning is proposed, in which the link recovery resources and the degree of importance of the damaged link in the failed service are taken into account to set the reward and penalty functions as well as the selection rules. Then the optimal link recovery combination is obtained through the accumulation of the maximum reward values. The simulation results show that the power communication network failure recovery algorithm proposed in this paper can restore quite considerable failed services quickly with limited resources.

Key words: smart grid, power communication network, large-scale fault, reinforcement learning, fault recover

贾惠彬, 盖永贺, 李保罡, 郑宏达. 基于强化学习的电力通信网故障恢复方法[J]. 中国电力, 2020, 53(6): 34-40.

JIA Huibin, GAI Yonghe, LI Baogang, ZHENG Hongda. Power Communication Network Recovery from Large-Scale Failures Based on Reinforcement Learning[J]. Electric Power, 2020, 53(6): 34-40.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.201907078

https://www.electricpower.com.cn/CN/Y2020/V53/I6/34

参考文献

[1] 谢迎军, 王玉亭, 李炜, 等. 电力企业通信网运行质量评估指标体系[J]. 中国电力, 2017, 50(10): 22-27
XIE Yingjun, WANG Yuting, LI Wei, et al. Indicator system for operation quality of electric power communication network[J]. Electric Power, 2017, 50(10): 22-27
[2] CHIARADONNA S, DI GIANDOMENICO F, MASETTI G. Analyzing the impact of failures in the electric power distribution grid[C]//2016 Seventh Latin-American Symposium on Dependable Computing (LADC). Cali, Colombia. IEEE, 2016: 99-108.
[3] LI H T, LI Q, JIANG Y, et al. A declarative failure recovery system in software defined networks[C]//2016 IEEE International Conference on Communications (ICC). Kuala Lumpur, Malaysia. IEEE, 2016.
[4] BERNSTEIN A, BIENSTOCK D, HAY D, et al. Power grid vulnerability to geographically correlated failures: Analysis and control implications[C]//IEEE INFOCOM 2014 - IEEE Conference on Computer Communications. Toronto, ON, Canada. IEEE, 2014: 2634-2642.
[5] HASELTINE C, EMAN EMANE S. Prediction of power grid failure using neural network learning[C]//2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA). Cancun, Mexico. IEEE, 2017: 505−510.
[6] FAN W L, LIU Z G, PING H, et al. Cascading failure model in power grids using the complex network theory[J]. IET Generation, Transmission & Distribution, 2016, 10(15): 3940-3949.
[7] 韩丽芳, 胡博文, 杨军, 等. 基于攻击预测的电力CPS安全风险评估[J]. 中国电力, 2019, 52(1): 48-56
HAN Lifang, HU Bowen, YANG Jun, et al. A new security risk assessment method for cyber physical power system based on attack prediction[J]. Electric Power, 2019, 52(1): 48-56
[8] 范君军, 陈智超, 陈涛, 等. 电网应急演练评估及其指标体系[J]. 中国电力, 2017, 50(11): 116-120
FAN Junjun, CHEN Zhichao, CHEN Tao, et al. Power grid emergency drill evaluation and the index system[J]. Electric Power, 2017, 50(11): 116-120
[9] 刁浩然, 杨明, 陈芳, 等. 基于强化学习理论的地区电网无功电压优化控制方法[J]. 电工技术学报, 2015, 30(12): 408-414
DIAO Haoran, YANG Ming, CHEN Fang, et al. Reactive power and voltage optimization control approach of the regional power grid based on reinforcement learning theory[J]. Transactions of China Electrotechnical Society, 2015, 30(12): 408-414
[10] MORADI M. A centralized reinforcement learning method for multi-agent job scheduling in Grid[C]//2016 6th International Conference on Computer and Knowledge Engineering (ICCKE). Mashhad, Iran. IEEE, 2016.
[11] YAN J, HE H B, ZHONG X N, et al. Q-learning-based vulnerability analysis of smart grid against sequential topology attacks[J]. IEEE Transactions on Information Forensics and Security, 2017, 12(1): 200-210.
[12] 汪洋, 高晗星, 周生平, 等. 融合拓扑及业务特性的电力通信网关键节点识别[J]. 中国电力, 2018, 51(10): 111-118
WANG Yang, GAO Hanxing, ZHOU Shengping, et al. Critical node identification for electric power communication network based on characteristics of topology and services[J]. Electric Power, 2018, 51(10): 111-118
[13] 陈国炎, 张哲, 尹项根, 等. 广域后备保护通信模式及其性能评估[J]. 中国电机工程学报, 2014, 34(1): 186-196
CHEN Guoyan, ZHANG Zhe, YIN Xianggen, et al. Wide area backup protection communication mode and its performance evaluation[J]. Proceedings of the CSEE, 2014, 34(1): 186-196
[14] 余才志. 必经节点的最短路径算法研究 [D]. 天津: 天津大学, 2016.
YU Caizhi. The research on the shortest path algorithm for the specified node[D]. Tianjin: Tianjin University, 2016.
[15] 康文雄, 许耀钊. 节点约束型最短路径的分层Dijkstra算法[J]. 华南理工大学学报(自然科学版), 2017, 45(1): 66-73
KANG Wenxiong, XU Yaozhao. A hierarchical dijkstra algorithm for solving shortest path from constrained nodes[J]. Journal of South China University of Technology(Natural Science Edition), 2017, 45(1): 66-73
[16] 杨春霞. 网络大规模毁坏后的渐进恢复机制研究[D]. 成都: 电子科技大学, 2012.
[17] PILEROOD A E, HEYDARI M, MAZDEH M M. A two-stage greedy heuristic for a flowshop scheduling problem under time-of-use electricity tariffs[J]. South African Journal of Industrial Engineering, 2018,29(1):143-154.
[18] 何瑞江, 胡志坚, 李燕, 等. 含分布式电源配电网故障区段定位的线性整数规划方法[J]. 电网技术, 2018, 42(11): 3684-3692
HE Ruijiang, HU Zhijian, LI Yan, et al. Fault section location method for DG-DNs based on integer linear programming[J]. Power System Technology, 2018, 42(11): 3684-3692
[19] 崔文岩, 孟相如, 杨欢欢, 等. QoS约束的链路故障多备份路径恢复算法[J]. 电子与信息学报, 2016, 38(8): 1850-1857
CUI Wenyan, MENG Xiangru, YANG Huanhuan, et al. Link failure recovery algorithm based on multiple backup paths with QoS constraint[J]. Journal of Electronics & Information Technology, 2016, 38(8): 1850-1857
[20] 陈晓玲, 杨军, 罗超, 等. 一种大电网潮流转移路径快速搜索方法[J]. 电网技术, 2015, 39(4): 1045-1052
CHEN Xiaoling, YANG Jun, LUO Chao, et al. A high-speed searching method for power flow transferring paths in large power grid[J]. Power System Technology, 2015, 39(4): 1045-1052

基于强化学习的电力通信网故障恢复方法

Power Communication Network Recovery from Large-Scale Failures Based on Reinforcement Learning

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	曾仪, 周毅, 陆继翔, 周良才, 唐宁恺, 李红. 基于多智能体安全深度强化学习的电压控制[J]. 中国电力, 2025, 58(2): 111-117.
[2]	孔志恒, 谭冲, 唐培耀, 胡成博, 郑敏. 基于Light-Resnet卷积神经网络的电力设备监测数值识别算法[J]. 中国电力, 2024, 57(8): 206-213.
[3]	焦昊, 殷岩岩, 吴晨, 刘建, 徐春雷, 徐贤, 孙国强. 基于安全强化学习的主动配电网有功-无功协调优化调度[J]. 中国电力, 2024, 57(3): 43-50.
[4]	李超英, 檀勤良. 基于智能体建模的新型电力系统下火电企业市场交易策略[J]. 中国电力, 2024, 57(2): 212-225.
[5]	何天玲, 蔡立波. 耦合电网业务约束条件的电力通信故障检修辅助决策的实现[J]. 中国电力, 2024, 57(12): 178-187.
[6]	薛松萍, 高德荃, 赵子岩, 林彧茜, 广泽晶, 张大卫. 基于业务差异化传输需求下的电力通信网路由算法[J]. 中国电力, 2024, 57(11): 183-190.
[7]	于宗超, 文明, 李湘华, 谢欣涛, 杨洪明. 融合群体智慧的分布式智能电网高效发展管理策略[J]. 中国电力, 2024, 57(10): 57-68.
[8]	张超, 赵冬梅, 季宇, 张颖. 基于改进深度Q网络的虚拟电厂实时优化调度[J]. 中国电力, 2024, 57(1): 91-100.
[9]	彭琳钰, 刘旭, 汤玮, 刘晴, 方浩, 张光辉. 基于对抗性强化学习的配电网融合通信效率提升方法[J]. 中国电力, 2023, 56(9): 127-133.
[10]	石文喆, 李冰洁, 尤培培, 张泠. 基于深度强化学习的建筑能源系统优化策略[J]. 中国电力, 2023, 56(6): 114-122.
[11]	宋伟业, 刘灵玥, 阎洁, 王航宇, 何书凯, 韩爽, 王明辉, 刘永前. 基于深度强化学习的海上风电集群自进化功率平滑控制方法[J]. 中国电力, 2023, 56(3): 36-46.
[12]	张淑兴, 马驰, 杨志学, 王尧, 吴昊, 任洲洋. 基于深度确定性策略梯度算法的风光储系统联合调度策略[J]. 中国电力, 2023, 56(2): 68-76.
[13]	吴岩, 王光政. 基于CiteSpace的配电网韧性评估与提升研究综述与展望[J]. 中国电力, 2023, 56(12): 100-112, 137.
[14]	李娟, 汪家铭, 许苏迪, 姜云龙, 杨雄. 基于量子萤火虫算法的配电网故障恢复策略[J]. 中国电力, 2023, 56(12): 191-198.
[15]	汤锦慧, 伍发元, 支妍力, 毛梦婷, 代小敏. 基于深度强化学习的户内变电站通风降噪优化设计[J]. 中国电力, 2023, 56(1): 96-105,118.

模态框（Modal）标题

基于强化学习的电力通信网故障恢复方法

Power Communication Network Recovery from Large-Scale Failures Based on Reinforcement Learning

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics