基于最小生成树及改进遗传算法的含分布式电源配电网孤岛划分方法

发布时间：2019-06-05 11:35

【摘要】：为在配电网大规模失电条件下充分发挥分布式电源作用,减小停电损失,基于分布式电源孤岛在线划分的思想,提出了配电网故障后分布式电源孤岛划分的原则,建立了分布式电源孤岛划分的数学模型。提出了孤岛划分算法中无向图的权值整定方法,在此基础上提出分为2个步骤的孤岛划分方法:第1步为基于Kruskal算法的含分布式电源最优辐射状配电网的生成方法;第2步为基于改进遗传算法的多DG配电网最优孤岛划分方法。基于PGE69节点系统对负荷权重设置对算法的影响进行了分析,并对各种方法进行了对比验证。结果表明,最优辐射状配电网是进行最优孤岛划分的基础,可缩小解集空间;所提方法相比启发式算法、分支定界法和动态规划法在恢复负荷数量、负荷价值方面具有优越性,可应用于配电网运行控制。
[Abstract]:In order to give full play to the role of distributed power supply and reduce the loss of power outage under the condition of large-scale power loss in distribution network, based on the idea of on-line division of distributed power islands, the principle of isolated islands division of distributed power supply after distribution network failure is put forward. The mathematical model of isolated island division of distributed power supply is established. In this paper, the weight setting method of undirected graph in isolated island partition algorithm is proposed, and a two-step islanding partition method is proposed: the first step is the generation method of optimal radiative distribution network with distributed power supply based on Kruskal algorithm; The second step is the optimal isolated island partition method for multi-DG distribution network based on improved genetic algorithm. Based on PGE69 node system, the influence of load weight setting on the algorithm is analyzed, and various methods are compared and verified. The results show that the optimal radiative distribution network is the basis of optimal islanding division, and the solution set space can be reduced. Compared with heuristic algorithm, branch and bound method and dynamic programming method have advantages in restoring load quantity and load value, and can be applied to distribution network operation control.
【作者单位】：中国电力科学研究院配电变压器节能技术北京市重点实验室;
【基金】：国家电网公司科技项目(PD71-12-010)~~
【分类号】：TM73

【参考文献】