Abstract: This paper proposes a two-stage selection scheme to deal with the problem that there may be several optimal solutions in the islanding partition of distribution system with distribution generations. This scheme utilizes the genetic algorithm to obtain the alternative population through the priority optimization; the optimal solution is selected from the alternative population based on the power quality index. The simulation implemented on American PG&E 69-node distribution system demonstrates that the possibility of generating the islands with low power quality is reduced by 45%. The case study shows that the two-stage islanding partition scheme proposed can decrease the possibility of generating the islands with low power quality significantly.

Key words: islanding partition, distributed energy resource, two-stage partition, power quality, genetic algorithm

摘要： 针对配电网的孤岛划分可能存在多个最优解的问题，提出了一套筛选最优解的两阶段方案。在第一阶段中，利用遗传算法通过优先级优化产生备选种群；第二阶段，在保证优先级的前提下，利用电能质量参数对备选种群进行筛选，选择出能形成具有较高电能质量的孤岛的最优解。在PG&E 69节点配电网上的仿真显示，产生具有较差电能质量的孤岛的可能性降低了45%。仿真结果说明提出的两阶段孤岛划分策略能够在保证优先级的前提下显著减少形成具有较低的电能质量的孤岛的可能性。

关键词: 孤岛划分, 分布式能量源, 两阶段划分, 电能质量, 遗传算法