极端灾害复电调度优化模型

Optimization Model for Power Restoration Scheduling in Extreme Disasters

摘要: 极端灾害给配电网带来多重不确定性，严重影响供电恢复。本文提出一种两阶段协同恢复方法：首先构建多不确定性模型，涵盖设备故障、负荷波动与分布式电源出力，建立关键设备故障概率模型及负荷动态响应模型；其次分析极端场景下配电网多模态运行特性与韧性瓶颈，构建灾前预防、灾中应急与灾后恢复的两阶段协同优化模型，目标为最小化负荷损失成本、最大化恢复节点数及最小化系统恢复时间，并考虑资源部署、功率平衡及抢修调度等约束。采用混合整数线性规划求解。在39节点系统上的仿真验证了该方法在提升复电效率与降低经济损失方面的有效性。

Abstract: Extreme disasters impose multiple uncertainties upon distribution networks. This paper proposes a two-stage collaborative restoration approach. Firstly, a multi-uncertainty model is constructed, encompassing equipment failures, load fluctuations, and distributed generation output. Probability models for critical equipment failures—such as lines and transformers—under disaster conditions are established, alongside dynamic response models for industrial, commercial, and residential loads. Secondly, the multi-modal operational characteristics and resilience bottlenecks of distribution networks under extreme scenarios are analysed, identifying critical issues during faults, supply-demand imbalances, and islanded operation states. Building upon this, a two-stage collaborative optimisation model is proposed, encompassing pre-disaster prevention, in-disaster emergency response, and post-disaster recovery. The model’s objectives are to minimise load loss costs, maximise the number of restored supply nodes, and minimise system restoration time, whilst accounting for constraints including resource deployment, power balancing, and repair scheduling. Finally, a solution method based on mixed-integer linear programming was designed. Its effectiveness in power restoration was demonstrated through simulation validation on a 39-node distribution network.