中文摘要：

分析了配电网的特点,提出使用改进的知识进化算法（Improved Knowledge Evolutionary Algorithm,IKEA）求解配电网重构问题。重点阐述了编码规则,设计了适合于配电网重构的传承操作、创新操作和检验操作,应用IKEA求解配电网重构问题。针对IEEE16、IEEE33和IEEE69节点的100次仿真实验中,IKEA全部找到最优解,平均收敛代数分别为9.7、10.2和18.1。同相关文献的智能算法求解配电网重构问题的测试结果对比发现,IKEA可有效地缩小搜索空间,减少迭代次数,而且解的效果较好,算法的效率较高。

英文摘要：

This paper puts forward a novel improved knowledge evolutionary algorithm for distribution network reconfiguration,which consists of a new coding rule,inheritance operation,innovation operation and inspection operation.In the 100 times simulation tests of IEEE16-bus,IEEE33-bus and IEEE69-bus distribution networks,IKEA can find the global optimal solution in each time,and the average convergence generations are 9.7,10.2 and 18.1 respectively.In the optimization of distribution network,IKEA reduces the times of iteration.The simulation results show that IKEA is feasible and efficient for distribution network reconfiguration.