CAJ | 학술논문

针对传统遗传算法在解决具有径向约束的配电网重构( DSR)问题时，搜索寻优难以得到准确的全局最优解，且采用二进制编码，使得编码位串长度较大等不足，提出了一种基于EWD( edge-window-decoder)编码的遗传算法。该遗传算法使用边窗( EWD)编码，将损耗最少及其他约束条件作为适应度函数，通过遗传操作，在所有满足约束条件的解(径向拓扑)中，选择一个最优的网络拓扑结构。通过对一典型的84节点配电系统进行重构举例，与其他方法进行比较，数据结果表明，提出的算法具有一定的实用性和有效性。
침대전통유전산법재해결구유경향약속적배전망중구( DSR)문제시，수색심우난이득도준학적전국최우해，차채용이진제편마，사득편마위천장도교대등불족，제출료일충기우EWD( edge-window-decoder)편마적유전산법。해유전산법사용변창( EWD)편마，장손모최소급기타약속조건작위괄응도함수，통과유전조작，재소유만족약속조건적해(경향탁복)중，선택일개최우적망락탁복결구。통과대일전형적84절점배전계통진행중구거례，여기타방법진행비교，수거결과표명，제출적산법구유일정적실용성화유효성。
Aiming at the problem of traditional genetic algorithm in solving distribution network reconfiguration ( DSR) with radial constraint that it is difficult to get the global optimal accurate solution, while using the binary cod-ing, the bit string is length and larger. A genetic algorithm method based on EWD( edge-window-decoder) is presen-ted, taking the least line loss and other constrains as functions to select an optimal solution from all that meets con-strain condition ( radial topology ) through genetic operation. Through the reconstruction examples of a typical 84 nodes in the distribution system, data results show that the proposed algorithm is practical and effective compared with other methods.