基于并行遗传算法的电力系统无功优化  

Reactive power optimization based on distributed genetic algorithm in electricpower system

作　　者：盛兆俊[1] 刘观起[1] 马燕峰[1] 

机构地区：[1]华北电力大学电力工程系,河北保定071003

出　　处：《华北电力大学学报（自然科学版）》2004年第1期11-14,共4页Journal of North China Electric Power University：Natural Science Edition

摘　　要：为了快速有效地求解大型电力系统无功优化的问题,提出了一种基于局域网的分布式并行遗传算法,采用主从式模型来组织局域网内的多台计算机进行并行计算,整个并行系统由一个主处理器和若干个从处理器组成。根据无功优化的特点,为了找到全局最优解,就目标函数及相关参数值的选取和遗传操作进行了改进。算例表明,该方法不仅有效地减少了系统的网络损耗,而且显著地提高了计算速度。In order to solve effectively the problem of optimizing reactive power of large scale power system, the paper provides a distributed collateral genetic algorithm based on the local area network. It adopts a principal and subordinate model to organize several computers to calculate synchronously. The whole system is made of a main processor and several slave processors. According to the characteristic of reactive power optimization and in order to find the overall optimal situation, the authors have improved the selection of goal functions and relevant parameter values as well as the heredity operation.

关 键 词：电力系统 无功优化 并行遗传算法 潮流计算 局域网 

分 类 号：TM744[电气工程—电力系统及自动化] TM769