Preventive maintenance is applied in distribution networks to prevent failures by performing maintenance actions on components that are at risk. Distributed generators (DGs) and batteries can be used to support power to nearby loads when they are isolated due to maintenance. In this paper, a novel short-term preventive maintenance method is proposed that explicitly considers the support potential of DGs and batteries as well as uncertainties in the power generated by the DGs. Two major issues are addressed. To deal with the large-scale complexity of the network, a depth-first-search clustering method is used to divide the network into zones. Moreover, a method is proposed to capture the influence of maintenance decisions in the model of the served load from DGs and batteries via generation of topological constraints. Then a stochastic scenario-based mixed-integer non-linear programming problem is formulated to determine the short-term maintenance schedule. We show the effectiveness and efficiency of the proposed approach via a case study based on a modified IEEE-34 bus distribution network, where we also compare a branch-and-bound and a particle swarm optimization solver. The results also show that the supporting potential of DGs and batteries in preventive maintenance scheduling allows a significant reduction of load losses.

中文翻译：

---

带有分布式发电机和电池的配电网络的短期预防性维护计划方法

预防性维护应用于配电网络中，以通过对有风险的组件执行维护操作来预防故障。分布式发电机（DG）和电池由于维护而被隔离时，可用于为附近的负载供电。本文提出了一种新颖的短期预防性维护方法，该方法明确考虑了DG和电池的支持潜力以及DG产生的电力的不确定性。解决了两个主要问题。为了应对网络的大规模复杂性，使用深度优先搜索聚类方法将网络划分为多个区域。此外，提出了一种方法，该方法通过生成拓扑约束来捕获服务决策中来自DG和电池的负载决策模型的影响。然后，提出了一种基于随机情景的混合整数非线性规划问题，用于确定短期维护计划。通过基于修改后的IEEE-34总线分配网络的案例研究，我们展示了该方法的有效性和效率。在该案例中，我们还比较了分支定界法和粒子群优化求解器。结果还表明，DG和电池在预防性维护计划中的支持潜力可以显着减少负载损失。