考虑数据中心负载灵活迁移特性的主动配电网阻塞管理

薛霖; 周越; 王建学; 古宸嘉; 侯果; 田子豪

doi:10.7652/xjtuxb202602019

您当前的位置：

首页 >

文章列表页 >

考虑数据中心负载灵活迁移特性的主动配电网阻塞管理

更新时间：2026-02-09

- 考虑数据中心负载灵活迁移特性的主动配电网阻塞管理
- Active Distribution Network Congestion Management Considering Flexible Migration Characteristics of Data Center Workload
- 西安交通大学学报 2026年60卷第2期页码：195-206
- 作者机构：
  
  西安交通大学电气工程学院，710049，西安
  西安交通大学电工材料电气绝缘全国重点实验室，710049，西安
  天津大学电气自动化与信息工程学院，300072，天津
  四川大学电气工程学院，610065，成都
- 作者简介：
  
  作者简介：薛霖（1997—），男，博士生；
  王建学（通信作者），男，教授，博士生导师。
- 基金信息：
  
  国家重点研发计划资助项目。(2022YFB2403500)
- DOI：10.7652/xjtuxb202602019
  中图分类号： TM72
- 收稿：2025-06-27，
  
  纸质出版：2026-02-10
- 稿件说明：
移动端阅览
薛霖, 周越, 王建学, 等. 考虑数据中心负载灵活迁移特性的主动配电网阻塞管理[J]. 西安交通大学学报, 2026,60(2):195-206.

XUE Lin, ZHOU Yue, WANG Jianxue, et al. Active Distribution Network Congestion Management Considering Flexible Migration Characteristics of Data Center Workload[J]. Journal of Xi'an Jiaotong University, 2026, 60(2): 195-206.
薛霖, 周越, 王建学, 等. 考虑数据中心负载灵活迁移特性的主动配电网阻塞管理[J]. 西安交通大学学报, 2026,60(2):195-206. DOI： 10.7652/xjtuxb202602019.

XUE Lin, ZHOU Yue, WANG Jianxue, et al. Active Distribution Network Congestion Management Considering Flexible Migration Characteristics of Data Center Workload[J]. Journal of Xi'an Jiaotong University, 2026, 60(2): 195-206. DOI： 10.7652/xjtuxb202602019.

摘要

针对新能源大量接入和数据中心负载分布不均导致的配电网阻塞问题，提出了基于模型-数据混合驱动的两阶段分层阻塞管理方法。在第1阶段，考虑主动配电网中有载调压变压器和并联电容器组的调节能力，建立了基于直接资源控制的主动配电网长时间尺度阻塞管理方法，并采用数据驱动的深度强化学习算法优化其投切档位。在第2阶段，建立了基于数据中心集群负载灵活迁移特性的主动配电网短时间尺度阻塞管理方法，将数据中心动态频率调节技术和动态服务器配置技术与工作负载灵活迁移特性相结合，并采用模型驱动的二阶锥规划算法，对新能源逆变器和工作负载迁移进行优化。改进的IEEE 33节点系统测试结果表明：所提两阶段阻塞管理方法可有效解决系统线路阻塞问题，测试日最大支路负载率由139.59%降至83.92%；与传统优化方法相比，所提方法将求解用时缩短至18.73 s，成本仅增加0.8%，具有更好的求解效率和鲁棒性，实现了数据中心集群与主动配电网的友好协同互动。

Abstract

To address the congestion problem in distribution networks caused by high penetration of renewable energy and uneven spatial distribution of data center workload，a two-stage hierarchical congestion management method based on a model-data hybrid-driven approach is proposed.In the first stage，considering the regulation capabilities of resources such as on-load tap changers and shunt capacitor banks in the active distribution network，a long-timescale congestion management method based on direct resource control is established and optimized using a data-driven deep reinforcement learning algorithm to adj ust their switching positions.In the second stage，a short-timescale congestion management method for the active distribution network is developed based on the flexible migration characteristics of data center cluster workload.This method integrates dynamic frequency regulation technology and dynamic server configuration technology of data centers with flexible workload migration characteristics，and employs a model-driven second-order cone programming algorithm to optimize renewable energy inverters and workload migration.Test results on a modified IEEE 33-node system show that the proposed two-stage congestion management method can effectively alleviate line congestion，reducing the maximum branch load ratio from 139.59% to 83.92% on the test day.Compared with traditional optimization methods，the proposed method shortens the solution time to 18.73 s with only a 0.8% cost increase，demonstrating better solution efficiency and robustness，and achieving friendly and coordinated interaction between data center clusters and the active distribution network.

关键词

Keywords

references

李鹏,钟瀚明,马红伟,等.基于深度强化学习的有源配电网多时间尺度源荷储协同优化调控[J].电工技术学报, 2025, 40(5):1487-1502.

LI Peng, ZHONG Hanming, MA Hongwei, et al. Multi-timescale optimal dispatch of source-load-storage coordination in active distribution network based on deep reinforcement learning[J].Transactions of China Electrotechnical Society, 2025, 40(5):1487-1502.

LIANG Yaran, LI Peng, SU Wen, et al.Development of green data center by configuring photovoltaic power generation and compressed air energy storage systems[J].Energy, 2024, 292:130516.

DING Zhaohao, XIE Liye, LU Ying, et al.Emissionaware stochastic resource planning scheme for data center microgrid considering batch workload scheduling and risk management[J].IEEE Transactions on Industry Applications, 2018, 54(6):5599-5608.

LU Yu, XIANG Yue, HUANG Yuan, et al.Deep reinforcement learning based optimal scheduling of active distribution system considering distributed generation, energy storage and flexible load[J].Energy, 2023, 271 :127087.

丛鹏伟,胡泽春.基于SOP和VSC的交直流混合配电网两阶段阻塞管理[J].供用电, 2020, 37(10):22-28.

CONG Pengwei, HU Zechun.Two-stage congestion management in AC/DC hybrid distribution network based on SOP and VSC[J].Distribution & Utilization, 2020, 37(10):22-28.

李鹏,张培深,王成山,等.基于智能软开关与市场机制的有源配电网阻塞管理方法[J].电力系统自动化, 2017, 41(20):9-16.

LI Peng, ZHANG Peishen, WANG Chengshan, et al. Congestion management method for active distribution networks based on soft open point and market mechanism[J].Automation of Electric Power Systems, 2017, 41(20):9-16.

王育飞,李可铭,陈强,等.基于市场机制与移动式储能的有源配电网阻塞管理方法[J].现代电力, 2024, 41(6):1138-1147.

WANG Yufei, LI Keming, CHEN Qiang, et al.Congestion management method for active distribution network based on market mechanism and mobile energy storage system[J].Modern Electric Power, 2024, 41(6):1138-1147.

XU Da, XIANG Shizhe, BAI Ziyi, et al. Optimal multi-energy portfolio towards zero carbon data center buildings in the presence of proactive demand response programs[J].Applied Energy, 2023, 350:121806.

周世博,周明,孙黎滢,等.激发多元灵活性的数据中心协同优化运行方法[J].电网技术, 2024, 48(11):4417-4426.

ZHOU Shibo, ZHOU Ming, SUN Liying, et al.Optimal synergistic operation method for data center by activating multiple flexibilities[J].Power System Technology, 2024, 48(11):4417-4426.

张锞,王旭,杨宏坤,等.数据中心集群灵活边界下电力系统分布鲁棒优化调度方法[J].电力系统自动化, 2024, 48(7):235-247.

ZHANG Ke, WANG Xu, YANG Hongkun, et al. Distributionally robust optimal scheduling method for power system under flexibility boundaries of data center clusters[J].Automation of Electric Power Systems, 2024, 48(7):235-247.

董雷,杨子民,乔骥,等.基于分层约束强化学习的综合能源多微网系统优化调度[J].电工技术学报, 2024, 39(5):1436-1453.

DONG Lei, YANG Zimin, QIAO Ji, et al.Optimal scheduling of integrated energy multi-microgrid system based on hierarchical constraint reinforcement learning[J].Transactions of China Electrotechnical Society, 2024, 39(5):1436-1453.

YANG Hongrong, XU Yinliang, GUO Qinglai.Dynamic incentive pricing on charging stations for realtime congestion management in distribution network:an adaptive model-based safe deep reinforcement learning method[J]. IEEE Transactions on Sustainable Energy, 2024, 15(2):1100-1113.

KHAN O G M, YOUSSEF A, SALAMA M, et al. Management of congestion in distribution networks utilizing demand side management and reinforcement learning[J].IEEE Systems Journal, 2023, 17 (3 ):4452-4463.

张剑,崔明建,姚潇毅,等.基于数据驱动与物理模型的主动配电网双时间尺度协调优化[J].电力系统自动化, 2023, 47(20):64-71.

ZHANG Jian, CUI Mingjian, YAO Xiaoyi, et al.Dual-timescale active and reactive power coordinated optimization for active distribution network based on datadriven and physical model[J].Automation of Electric Power Systems, 2023, 47(20):64-71.

王志杨,张靖,何宇,等.数据与模型混合驱动的区域综合能源系统双层优化调度决策方法[J].电网技术, 2022, 46(10):3797-3809.

WANG Zhiyang, ZHANG Jing, HE Yu, et al.Hybrid data-driven and model-driven bi-level optimal scheduling decision for regional integrated energy systems[J].Power System Technology, 2022, 46(10):3797-3809.

张玉莹,曾博,周吟雨,等.碳减排驱动下的数据中心与配电网交互式集成规划研究[J].电工技术学报, 2023 , 38(23):6433-6450.

ZHANG Yuying, ZENG Bo, ZHOU Yinyu, et al. Research on interactive integration planning of data centers and distribution network driven by carbon emission reduction[J].Transactions of China Electrotechnical Society, 2023, 38(23):6433-6450.

JIN Chaoqiang, BAI Xuelian, YANG Chao, et al.A review of power consumption models of servers in data centers[J].Applied Energy, 2020, 265:114806.

CHO J, PARK B, JANG S.Development of an independent modular air containment system for high-density data centers:experimental investigation of rowbased cooling performance andPUE[J]. Energy, 2022, 258:124787.

王天琪,于浩,赵金利,等.算力-热力灵活性协同的数据中心能量管理方法[J].高电压技术, 2024, 50(9):4069-4079.

WANG Tianqi, YU Hao, ZHAO Jinli, et al.Optimal energy management of data centers considering the synergy of computing power and thermal power flexibility[J].High Voltage Engineering, 2024, 50 (9):4069-4079.

LIU Yang, LEI Shunbo, HOU Yunhe.Restoration of power distribution systems with multiple data centers as critical loads[J]. IEEE Transactions on Smart Grid, 2019, 10(5):5294-5307.

SUN Gang, ANAND V, LIAO Dan, et al.Power-efficient provisioning for online virtual network requests in cloud-based data centers[J].IEEE Systems Journal, 2015, 9(2):427-441.

张剑,崔明建,何怡刚.结合数据驱动与物理模型的主动配电网双时间尺度电压协调优化控制[J].电工技术学报, 2024, 39(5):1327-1339.

ZHANG Jian, CUI Mingjian, HE Yigang.Dual timescales coordinated and optimal voltages control in distribution systems using data-driven and physical optimization[J].Transactions of China Electrotechnical Society, 2024, 39(5):1327-1339.

XU Xiaoyuan, LI Yunhong, YAN Zheng, et al.Hierarchical central-local inverter-based voltage control in distribution networks considering stochastic PV power admissible range[J]. IEEE Transactions on Smart Grid, 2023, 14(3):1868-1879.

李鹏,姜磊,王加浩,等.基于深度强化学习的新能源配电网双时间尺度无功电压优化[J].中国电机工程学报, 2023, 43(16):6255-6265.

LI Peng, JIANG Lei, WANG Jiahao, et al.Optimization of dual-time scale reactive voltage for distribution network with renewable energy based on deep reinforcement learning[J].Proceedings of the CSEE, 2023, 43(16):6255-6265.

马庆,邓长虹.基于单/多智能体简化强化学习的电力系统无功电压控制[J].电工技术学报, 2024, 39(5):1300-1312.

MA Qing, DENG Changhong.Single/multi agent simplified deep reinforcement learning based volt-var control of power system[J].Transactions of China Electrotechnical Society, 2024, 39(5):1300-1312.

HAN G, JOO H J, LIM H W, et al.Data-driven heat pump operation strategy using rainbow deep reinforcement learning for significant reduction of electricity cost[J].Energy, 2023, 270:126913.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

数据中心空气冷却和风-液混合冷却的能耗比较

单相浸没冷却数据中心的多尺度模型研究

数据中心多能互补分布式能源系统设计与运行优化研究

数据中心空气冷却和风-液混合冷却的能耗比较

主动配电系统与主网的有功协调