跟随策略蜂群算法在电力经济调度中的研究  

Artificial bee colony algorithm with a following strategy and its application in economic dispatch

作　　者：宋锦[1,2] 高浩[2] 王保云[2] 

机构地区：[1]南京信息职业技术学院通信学院,江苏南京210023 [2]南京邮电大学自动化学院,江苏南京210023

出　　处：《南京邮电大学学报（自然科学版）》2016年第4期89-94,共6页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家自然科学基金(61571236);江苏省高校品牌专业建设工程(PPZY2015A092);中国博士后基金(2014M551632);江苏省博士后基金(1402018A)资助项目

摘　　要：蜂群算法自提出以来由于其良好的全局搜索能力一直受到广泛的注意,但相对于其他群体算法来看,传统的蜂群算法存在着收敛速度慢的缺陷。文中借鉴粒子群算法的迭代公式,在跟随蜂迭代公式引入一种新的策略,提出了一种新的基于gbest更新策略蜂群算法。当群体的迭代度达到设定的条件时,对跟随蜂使用一种全新的迭代公式,使它能够以一个整体获得更多的在gbest附近进行局部搜索的机会。标准测试函数的仿真结果表明,该算法在有效保证全局搜索能力的前提条件下,提升了算法的是收敛速度。针对电力系统的动态经济调度问题,改进以后的蜂群算法获得了较优的结果。Many researchers pay attention to the artificial bee colony（ ABC） algorithm for its power global search ability. Compared with other evolutionary algorithms,the conventional bee colony algorithm shows a slower convergence rate. Aimed at the merits of particle swarm Optimization（ PSO）,a new update strategy is introduced for the phase of onlook bee and a new ABC algorithm based on a gbest update strategy is presented. When the number of the current iteration achieves a defined condition,a new update equation is used for onlook bees. Then,they could have more opportunities to search the area nearby the gbest. Experimental results on the traditional benchmark functions show that algorithm can accelerates the convergence rate under the condition of guaranteeing its global search ability. The improved algorithm can obtain favorable results on the dynamic economic dispatch problem.

关 键 词：蜂群算法 动态经济调度 全局搜索 个体历史最优 收敛速度 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]