网络重构中基于配电网拓扑特征的改进遗传算法

Improvement of genetic algorithms in network reconfiguration based on topology characteristics

网络重构是配电系统运行和控制的重要手段,利用遗传算法求解配电网络重构问题可以较好地利用其智能计算的特点,但也存在计算速度慢与收敛于非可行解等问题。结合配电网络拓扑结构特征和遗传算法特点具体探讨了如何实现遗传算法的网络重构,对遗传操作中的选择、交叉和变异提出了基于配电网络拓扑的改进方法,提高了算法的收敛速度和收敛性,算例验证了所提方法的可行性和正确性。

Network reconfiguration is an important approach to distribution system operation and control. The genetic algorithm （GA） has some advantages such as intelligent computing in solving the distribution network reconfiguration problem, however, it also has the drawbacks in slow convergence and non-feasible solutions. Based on the characteristics of distribution network topology and that of the GA method, the issue of how to accomplish network reconfigura- tion based on GA method was discussed, and the strategy of population initialization, selection, crossover, and muta- tion was improved, which increased the convergence and the calculating speed. Example calculation proved the method feasible and correct.