基于改进小生境遗传算法的电力系统无功优化  

Reactive Power Optimization of Power System Based on Improved Niche Genetic Algorithm

作　　者：崔挺[1] 孙元章[1] 徐箭[1] 黄磊[1] 

机构地区：[1]武汉大学电气工程学院,湖北省武汉市430072

出　　处：《中国电机工程学报》2011年第19期43-50,共8页Proceedings of the CSEE

基　　金："十一五"国家科技支撑计划重大项目(2008BAA13B04);国家自然科学基金项目(51007067);中央高校基本科研业务费专项资金项目(5082008);国家电网公司科技项目(豫电调通KJ(2010)72号)~~

摘　　要：针对电力系统无功优化问题,提出一种改进小生境遗传算法来克服小生境遗传算法中小生境难以确定的不足,改善遗传算法容易陷入局部收敛和早熟的缺点。通过模糊动态聚类分析方法实现小生境群体的划分,然后利用适应度共享技术对小生境内个体适应度进行调整,以提高全局寻优能力。提出和运用隔代小生境共享机制、最优个体邻域搜索及保留策略等以提高算法的计算速度和收敛速度。通过对IEEE 57节点测试系统进行无功优化计算及结果分析,说明所提出算法的全局搜索能力强、效率高,能得到较好的结果。An improved niche genetic algorithm for reactive power optimization of power system was proposed to avoid the disadvantage of niche genetic algorithm that is difficult in determining the niches and improve the premature phenomenon and local convergence of genetic algorithm. By the method of fuzzy dynamic clustering, niche groups were created. And then a fitness sharing mechanism was applied to adjust individuals＇ fitness in each niche so as to enhance the global searching ability. In order to improve the convergence rate further more, mechanism of applying niche fitness sharing in every some generations, local searching to the elitist and retention strategy and other strategies were introduced into the algorithm. The results of calculating on IEEE 57-bus system show that it performs better in celerity, accuracy and efficiency.

关 键 词：电力系统 无功优化 遗传算法 小生境 模糊动态聚类 适应度共享 

分 类 号：TM74[电气工程—电力系统及自动化]