您的位置：山东大学 -> 科技期刊社 -> 《山东大学学报（理学版）》

《山东大学学报(理学版)》 ›› 2025, Vol. 60 ›› Issue (9): 41-51.doi: 10.6040/j.issn.1671-9352.0.2024.377

• • 上一篇    

基于长短时序预测的拓扑构建与控制

严莉1,呼海林1,王高洲1,张闻彬1,潘法定1,张啸2,郑艳伟2*   

  1. 1.国网山东省电力公司信息通信公司, 山东 济南 250013;2.山东大学计算机科学与技术学院, 山东 青岛 266237
  • 发布日期:2025-09-10
  • 通讯作者: 郑艳伟(1977— ),男,副研究员,博士,研究方向为时序数据预测、计算机视觉. E-mail:zhengyw@sdu.edu.cn
  • 作者简介:严莉(1975— ),女,高级工程师,硕士,研究方向为电力信息技术. E-mail:yanli@sd.sgcc.com.cn*通信作者:郑艳伟(1977— ),男,副研究员,博士,研究方向为时序数据预测、计算机视觉. E-mail:zhengyw@sdu.edu.cn
  • 基金资助:
    国网山东省电力公司科技资助项目(52062723000B)

Topology construction and control based on long short-term prediction

YAN Li1, HU Hailin1, WANG Gaozhou1, ZHANG Wenbin1, PAN Fading1, ZHANG Xiao2, ZHENG Yanwei2*   

  1. 1. State Grid Shandong Electric Power Company Information and Communication Company, Jinan 250013, Shandong, China;
    2. School of Computer Science and Technology, Shandong University, Qingdao 266237, Shandong, China
  • Published:2025-09-10

PDF (PC)

3

摘要： 为优化动态网络的拓扑构建与资源分配,提出基于长短时序预测的拓扑构建与控制(long short-term prediction-based topology construction and control, LSPTCC)框架。采用长短期记忆(long short-term memory, LSTM)网络和Informer模型进行多维时间序列的长时和短时预测,精准捕捉数据中的时间依赖性与非平稳性波动。基于预测结果,使用增强容量约束设计(enhanced capacity constrained design, ECCD)算法构建最小生成树(minimum spanning tree, MST),优化节点间的连接,减少传输路径的总损耗。利用最大网络流算法实现动态的流量分配与调整,确保系统在流量波动情况下的高效流量资源利用。实验采用光伏消纳数据集,结果表明该框架能够准确预测发电量和用电量,并通过优化拓扑结构和资源分配,减少电力传输损耗,验证LSPTCC框架的有效性和鲁棒性。

关键词: 长短时序预测, 最小生成树, 最大网络流算法, 动态网络拓扑

Abstract: To optimize dynamic network topology construction and resource allocation, a long short-term prediction-based topology construction and control(LSPTCC)framework is proposed. Long short-term memory(LSTM)network and Informer models are employed for long short-term prediction of multivariate time series. Temporal dependencies and non-stationary fluctuations in the data are accurately captured. Based on the prediction results, the enhanced capacity constrained design(ECCD)algorithm is used to construct a minimum spanning tree(MST)and to optimize the connections between nodes to minimize overall transmission losses. Additionally, a maximum network flow algorithm is applied to dynamically adjust resource allocation, ensuring efficient utilization of resources under fluctuating conditions. Experiments conducted on a photovoltaic consumption dataset demonstrate that the proposed framework can accurately predict power generation and consumption. The power transmission losses are effectively reduced by optimizing both the topology and resource allocation. The efficiency and robustness of the proposed algorithms are validated.

Key words: long short-term prediction, minimum spanning tree, maximum network flow algorithm, dynamic network topology

中图分类号: 

  • TP183
[1] YU Yong, SI Xiaosheng, HU Changhua, et al. A review of recurrent neural networks: LSTM cells and network architectures[J]. Neural Computation, 2019, 31(7):1235-1270.
[2] DEY R, SALEM F M. Gate-variants of gated recurrent unit(GRU)neural networks[C] //2017 IEEE 60th International Midwest Symposium on Circuits and Systems(MWSCAS). Medford: IEEE, 2017:1597-1600.
[3] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C] //Advances in Neural Information Processing Systems. Vancouver: Curran Associates, 2017:6000-6010.
[4] HAN Kai, WANG Yifan, CHEN Huazhong, et al. A survey on vision transformer[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 45(1):87-110.
[5] ZHOU Haibo, ZHANG Shuang, PENG Jianxin, et al. Informer: beyond efficient transformer for long sequence time-series forecasting[C] //Proceedings of the 35th AAAI Conference on Artificial Intelligence. Virtual: AAAI, 2021:11106-11115.
[6] 王红瑞,刘晓红,唐奇,等. 基于小波变换的支持向量机水文过程预测[J]. 清华大学学报(自然科学版),2010,50(9):1378-1382. WANG Hongrui, LIU Xiaohong, TANG Qi, et al. Hydrological process prediction based on wavelet transform and support vector machine[J]. Journal of Tsinghua University(Science and Technology), 2010, 50(9):1378-1382.
[7] 李洁,林永峰,陈亮,等. 基于深度经验模态分解的金融市场时序预测[J]. 计算机工程与设计,2019,40(12):3613-3619. LI Jie, LIN Yongfeng, CHEN Liang, et al. Financial market time series prediction based on deep empirical mode decomposition[J]. Computer Engineering and Design, 2019, 40(12):3613-3619.
[8] PAVLYSHENKO B. Using stacking approaches for machine learning models[C] //2018 IEEE Second International Conference on Data Stream Mining & Processing(DSMP). Lviv: IEEE, 2018:255-258.
[9] OSIPOV V, SANDERS P SINGLER J. The filter-kruskal minimum spanning tree algorithm[C] //The Eleventh Workshop on Algorithm Engineering and Experiments(ALENEX). New York: Society for Industrial and Applied Mathematics, 2009:52-61.
[10] LI Haiming, XIA Qiyang, WANG Yong. Research and improvement of Kruskal algorithm[J]. Journal of Computer and Communications, 2017, 5(12):63-69.
[11] MANEN S, GUILLAUMIN M, GOOL L V. Prime object proposals with randomized Prims algorithm[C] //Proceedings of the IEEE International Conference on Computer Vision(ICCV). Sydney: IEEE, 2013:2536-2543.
[12] DAGAR Sunny. Modified Prims algorithm[J]. Journal of Computer and Information Technology, 2012, 3(2):26-29.
[13] 来卫国,李鸥,程军,等. 一种新的求解度约束最小生成树的遗传算法[J]. 计算机仿真,2008,25(8):162-165. LAI Weiguo, LI Ou, CHENG Jun, et al. A new genetic algorithm for solving degree-constrained minimum spanning tree[J]. Computer Simulation, 2008, 25(8):162-165.
[14] 周毅,周良才,丁佳立,等. 基于深度强化学习的电网拓扑优化和潮流控制的方法[J]. 上海交通大学学报,2021,55(增刊2):25-31. ZHOU Yi, ZHOU Liangcai, DING Jiali, et al. Method for topology optimization and power flow control based on deep reinforcement learning[J]. Journal of Shanghai Jiao Tong University, 2021,55(Suppl.2):25-31.
[15] 富坤,高金辉,赵晓梦,等. 融合全局结构信息的拓扑优化图卷积网络[J]. 计算机应用,2022,42(2):357-364. FU Kun, GAO Jinhui, ZHAO Xiaomeng, et al. Topology optimization graph convolutional network integrating global structural information[J]. Computer Applications, 2022, 42(2):357-364.
[16] CORMEN T H, LEISERSON C E, RIVEST R L, et al. Introduction to algorithms[M]. Cambridge: MIT Press, 2009:714-731.
[17] EDMONDS J, KARP R M, Theoretical improvements in algorithmic efficiency for network flows[J]. Journal of the ACM, 1972, 19(2):248-264.
[18] GOLDBERG A V, TARJAN R E. A new approach to the maximum flow problem[J]. Journal of the ACM, 1988, 35(4):921-940.
[19] KORTE B, VYGEN J. Combinatorial optimization: theory and algorithms[M]. Berlin: Springer, 2012:177-214.
[20] ORLIN J B. Max flows in O(nm)time or better[C] //Proceedings of the 45th Annual ACM Symposium on Theory of Computing(STOC). Palo Alto: ACM, 2013:765-774.
[21] BIRGE J R, LOUVEAUX F. Introduction to stochastic programming[M]. New York: Springer, 2011.
[22] BEN-TAL A, NEMIROVSKI A. Robust optimization: methodology and applications[J]. Mathematical Programming, 2000,92(3):453-480.
[23] MIETTINEN K. Nonlinear multiobjective optimization[M]. Dordrecht: Kluwer Academic Publishers, 1999:239-249.
[24] DEAN J, GHEMAWAT S. Mapreduce: simplified data processing on large clusters[J]. Communications of the ACM, 2008, 51(1):107-113.
[1] 韦修喜,彭茂松,黄华娟. 基于改进蝴蝶算法的水文地质参数优化[J]. 《山东大学学报(理学版)》, 2024, 59(3): 37-50.
[2] 许侃,刘瑞鑫,林鸿飞,刘海峰,冯娇娇,李家平,林原,徐博. 基于异质网络嵌入的学术论文推荐方法[J]. 《山东大学学报(理学版)》, 2020, 55(11): 35-45.
[3] 张超,梁英,方浩汕. 支持隐私保护的社交网络信息推荐方法[J]. 《山东大学学报(理学版)》, 2020, 55(3): 9-18.
[4] 谢小杰,梁英,董祥祥. 社交网络用户敏感属性迭代识别方法[J]. 《山东大学学报(理学版)》, 2019, 54(3): 10-17, 27.
[5] 肖炜茗,王贵君. 基于Bernstein多项式的SISO三层前向神经网络的设计与逼近[J]. 山东大学学报（理学版）, 2018, 53(9): 55-61.
[6] 索春凤, 王贵君. 最大交互数对非齐次T-S模糊系统的潜在影响[J]. 山东大学学报（理学版）, 2015, 50(08): 14-19.
[7] 吴瑞海 董吉文 段琪庆. 变尺度混沌粒子群与小波的地基沉降预测应用[J]. J4, 2009, 44(11): 75-78.
[8] . 一种基于驾驶员学习过程的在线估计算法[J]. J4, 2009, 44(7): 83-88.
[9] 朱世伟,赛 英 . 基于主成分分析和粗径向基神经网络的财务预警模型研究[J]. J4, 2008, 43(11): 48-53 .
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
版权所有 © 2018 《山东大学学报(理学版)》编辑部　
地址: 山东省济南市山大南路27号山东大学中心校区（250100）　电话: +86-0531-88366917　E-mail: xblxb@sdu.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发