雷傲宇, 吴睿, 李家璐,等.面向蓄意破坏的电力系统重要负荷保供与自愈恢复策略[J].河南理工大学学报（自然科学版）,2026,45(5):97-106.

LEI A Y, WU R, LI J L, et al.Supply assurance and self-healing restoration strategies for critical loads in power systems against deliberate attacks[J].Journal of Henan Polytechnic University(Natural Science) ,2026,45(5):97-106.

面向蓄意破坏的电力系统重要负荷保供与自愈恢复策略

雷傲宇1, 吴睿2, 李家璐1, 李泽林2, 高永强1, 钟兴黎3, 周坤4, 罗胜耀2

1.中国南方电网 电力调度控制中心，广东 广州  510663；2.国电南瑞南京控制系统有限公司，江苏 南京  211106；3.昭通供电局 电力调度控制中心，云南 昭通  657000；4.大理供电局 电力调度控制中心，云南 大理  671000

摘要: 目的针对蓄意破坏触发的N-3级极端故障下电力系统自愈恢复与重要负荷保供挑战，以及传统恢复策略难以同时兼顾故障场景构造、重要负荷保护和减载公平性问题，构建一种面向关键负荷的自愈恢复策略框架。  方法首先，在交流潮流模型基础上引入边介数中心性与节点介数中心性等拓扑指标筛选关键输电线路，并施加多条线路同时断开的N-3组合，形成覆盖测试系统的极端故障场景；其次，构造拓扑-负荷融合的母线重要度，实现对节点结构关键性和负荷水平的统一量化；最后，围绕该重要度指标，构建统一框架下的多策略迭代负荷削减方法。基于4类基准减载策略，提出集成低压优先队列、低压微减载和发电机电压微提升的低压友好-关键负荷优先策略，并结合供电率变化与越限改善程度设计自适应步长调节与有限步终止机制，实现减载规模的按需调节。 结果仿真结果表明，与统一减载策略相比，所提策略可将总减载功率由812.1 MW降至54.9 MW，最终供电率由35.1%提高到95.6%，同时降低严重过载线路比例和低电压母线比例，此外，关键负荷削减指数和高重要度节点减载占比均明显降低。  结论自愈恢复策略不仅提高了供电保障水平和超限收敛速度，而且兼顾了关键负荷的保护和减载分配的公平性。

关键词:蓄意破坏;自愈恢复;负荷削减;重要负荷保护;减载公平性

doi:10.16186/j.cnki.1673-9787.2025120087

基金项目:中国南方电网公司科技项目（000005KC24010023）；国家自然科学基金资助项目（62203395）

收稿日期:2025/12/27

修回日期:2026/05/09

出版日期:2026-07-24

Supply assurance and self-healing restoration strategies for critical loads in power systems against deliberate attacks

Lei Aoyu1, Wu Rui2, Li Jialu1, Li Zelin2, Gao Yongqiang1, Zhong Xingli3, Zhou Kun4, Luo Shengyao2

1.Power Dispatching Control Center of China Southern Power Grid， Guangzhou  510663， Guangdong， China；2.NARI Technology Nanjing Control System Co.， Ltd.， Nanjing  211106， Jiangsu， China；3.Zhaotong Dispatch Control Center， Yunnan Power Grid Co.， Ltd.， Zhaotong  657000， Yunnan， China；4.Dali Dispatch Control Center， Yunnan Power Grid Co.， Ltd.， Dali  671000， Yunnan， China

Abstract: Objectives The challenges of power system self-healing restoration and critical load supply assurance under N-3 level extreme contingencies triggered by deliberate attacks were addressed. A self-healing restoration strategy framework oriented toward critical loads was constructed to address the difficulty of conventional restoration strategies in simultaneously considering contingency construction， critical-load protection， and load-shedding fairness.  Methods First， an AC power flow model was used as the basis. Critical transmission lines were identified through topological metrics， including edge betweenness centrality and node betweenness centrality. Multiple simultaneous line outages were applied， forming N-3 combinations to generate extreme contingency scenarios covering the test system. Second， a bus importance metric was constructed by integrating topology and load， achieving unified quantification of nodal structural criticality and load level. Finally， a multi-strategy iterative load shedding method was constructed within a unified framework around the proposed importance metric. Based on four baseline load-shedding strategies， a low-voltage-friendly critical-load-priority strategy was proposed， integrating a low-voltage priority queue， low-voltage micro-shedding， and generator voltage micro-boosting. An adaptive step-size adjustment and a finite-step termination mechanism were designed according to supply-rate variation and violation-improvement performance， enabling on-demand regulation of the shedding scale.  Results Simulation results show that， compared with the uniform load-shedding strategy， the proposed strategy reduces the total shed power from 812.1 MW to 54.9 MW and improves the final supply rate from 35.1% to 95.6%， while significantly reducing the proportions of overloaded lines and low-voltage buses. In addition， the critical load shedding index and the shedding proportion of high-importance buses were both reduced.  Conclusions The proposed self-healing restoration strategy enhances power-supply assurance and speeds up convergence of violation-elimination， while balancing critical-load protection with fairness in load-shedding allocation.

Key words:deliberate attacks;self-healing restoration;load shedding;critical load protection;load shedding fairness