基于改进遗传算法配电网网架规划及软件开发.pdf

Abstract 摘要遗传算法是一种崮适应全局优化概率搜索方法。其不依赖于问题的具体领域且具有很高的鲁棒性,近年在配电网网架规划中已有大量深入的研究。但在具体应用过程中仍存在规划前期准备工作多,规划过程难以反应规划问题本质、寻优过程与配电网规划信息结合不够充分等问题。本文在现有{舞究成果基礁上,提滋一种基于待规划负荷节点数的定长度实数编码方式,并采用基于到已建网架区域的短距离馈线集合为核心的解码方式。解码原则为就近选择。遗传算法方面根据CHc策略加以改进,遗传操作过程采用了考虑变量微调的改进片段重组交叉算子和改进置换算子。为保证算法寻优的前期稳定性,变异算子只在迭代后期参与操作。整个过程考虑了解码过程和配电网网架方案形成特点,形成的方案均可行合理。而具有最高选择压的(“+九)选择算子为方案寻优保证了进化收敛速度。此外,根据闭环设计、开环运行的原则,本算法在最优运行辐射型网絮结构基础上对站阀、站内联络接线和联络开关进行优化设置,得到了最终手拉手配电网网架结构方案的费用。最后对最优网架的供电可靠性进行校验并得到故障情况联络接线选择预案。在上述理论基础上,本文进行了基于MFC的配电网网架结构规划程序编写和基于Arc嚣ngine的规划GIs软件的二次开发,荠最终实现规划功能和规划方案创建与可视化,空间地理信息存储,线路潮流渲染等功能。综上所述,规划方案形成过程无需获取待选线路集,能准确反映网架规划问题实痰。规划算铡的寻优结果表明了基于改进遗传算法的配电阙网架规划方案趋实用性和可行性。关键词:配电网规划;网架规划;改进遗传算法;软件开发 Abstract Ab stract ic Algorithm(GA)is an adaptiVe and probabilistic search method of global itdoes not rely on thespecific area oftheproblem and has high robustness,the research about GA es in-depth inpower system planning in recent years. HoweVer,in engineering applications,the distribution work deVeloped by GA hassome problems,such asheaVy preparing work inprimary ,it’s dif右cultinreflecting the nature ofprogramming problems bining adequately theinfbmation intheplanning procedure. According tothecurrent research result,the thesispresented anew real一number encoding based on thenumbers ofload nodes inplanning area and short f色eder lines collection as the essence to execute work planning automation. Relying on the improVed CHC strategy,the ic operation adopted improVed Fragment Reordering CrossoVer operator considering the shift operator and Swap Mutation the stability oftheoptiInizing process,the mutation took part inthe ic operation just inthe wh01e process blending the decoding and the characteristic of work’s stmcture ensured each solution eHicient and the(p+九)selector thatowns highest selectiVe pressure guaranteed the addition,basing on theprinciple of open一100p operation and closed一100p design,the