Focusing on the problem of premature convergence and slow convergence of the standard genetic algorithm, this paper proposes a hybrid genetic algorithm (adaptive genetic annealing algorithm) to solve the fault location in radialized distribution networks. This algorithm adopts the combined mechanism of the roulette strategy and the optimal strategy to keep the current best individual in the population, and uses the adaptive crossover and mutation probability to expand the search area about population, and then introduces simulated annealing algorithm, so as to speed up the convergence rate of the interactive post. Finally, a simulation calculation is conducted for the IEEE-33 system and the result indicates that the algorithm has fast convergent velocity and accurate fault location abilities in single fault or multiple faults. In addition, it also has good fault-tolerance when fault information is aberrated.%针对标准遗传算法易早熟收敛以及收敛速度慢的问题,提出了一种混合遗传算法(自适应遗传退火算法)用于解决辐射状配电网故障定位问题.该算法采用轮盘赌和最优保存策略相结合的选择机制,使得当前最优个体始终保持在种群里,并结合自适应交叉、变异概率,扩大种群的搜索范围,继而引入模拟退火算法,加快迭代后期算法的收敛速度.最后,通过对IEEE-33节点配电系统进行仿真计算,结果表明,该算法能够对单点和多点故障进行实时、准确地定位,并在故障信息畸变的情况下,也能快速地得到准确结果.
展开▼
