一种适用于配电网重构的动态调整策略

摘要

提出了一种适用于配电网重构的动态调整策略,将其与整数粒子群优化算法相结合,实现了高效的配电网重构.动态调整策略分为问题变量解空间动态调整策略和粒子位置变量截断方式动态调整策略两部分:前者是指初始化时,粒子位置向量中的每一维变量根据已经更新的问题变量动态调整问题解空间,从而保证映射的问题解都是拓扑可行解；后者是指粒子更新过程中,位置向量的每一维变量通过动态调整截断方式来改变粒子位置向量空间到问题解空间的映射关系,潜在的拓扑不可行解被视为截断误差偏大而自动优化调整,这样不仅使更新过程不产生拓扑不可行解,而且更新过程的优化特性得以保留.动态调整策略可以结合粒子连续运动空间和问题离散解空间的映射关系,避免粒子搜索过程中产生网络拓扑不可行解,不仅达到了减少冗余搜索的目的,还提高了算法的优化性能.算例结果表明,提出的结合整数粒子群映射关系的动态调整策略不仅能够有效地求解配电网重构问题,而且具有良好的鲁棒性.%A dynamic adjustment strategy (DAS) for distribution network reconfiguration is proposed, and successfully combined with integer particle swarm optimization (IPSO) algorithm to solve distribution network reconfiguration problem efficiently. DAS can be divided into two steps - one refers to dynamic adjustment for solution space of variables of reconfiguration and the other for round method of locator variables of particles. The former guarantees the initialized reconfiguration solution satisfying the topology constraints after solution space of variables of reconfiguration is adjusted dynamically according to the updated variables. The latter ensures the reconfiguration solution in every update procedure to get feasible for topology due to dynamic adjustment for round method of locator variables of particles, and the potential infeasible solution is auto-adjusted optimally. DAS combines the mapping from continuous space for particle to discrete space for reconfiguration to avoid reconfiguration solution infeasible for topology, thus decreases unnecessary search and improves optimization performance. The test systems reveal that the proposed method enables to effectively to solve distribution network reconfiguration with robustness.