基于自适应遗传退火算法的配电网孤岛划分方法

Islanding Method of Distribution System Based on Adaptive Genetic Annealing Algorithm

含分布式电源的配电网发生故障后,孤岛运行可以保障重要负荷的供电。根据配电网结构和负荷可控性,将配电网简化为节点赋权的连通图,以恢复的重要负荷最多为主要目标,建立孤岛划分问题模型。对支路进行0-1编码,结合节点-支路关联矩阵和广度优先搜索对个体进行解码,校验各区域是否满足约束要求,计算个体适应度,并对不可行解施以惩罚,采用自适应遗传退火算法寻优来划分孤岛。算例分析表明,所提出的方法能迅速制定优化的孤岛方案,验证了其有效性。

After a fault occurs, island operation can be used to maintain the power supply of critical loads in distribu- tion system with distributed generations （DGs）. According to the distribution structure and the controllability of loads, an islanding model is proposed by simplifying the distribution network as connected graph with weighted nodes and tak- ing the maximum priority load as the main objective. Each branch of distribution network is coded as 0-1, and node- branch incidence matrix and breadth-first search are adopted together to decode the islanding scheme. After the identification of all the areas, the fitness of a solution will be calculated while the infeasible solution should be punished. Adaptive genetic annealing algorithm is used to divide islands for distribution network. The simulation results demonstrate that the proposed algorithm can quickly figure out an optimal islanding solution, which proves its effectiveness.