摘要
提出一种遗传和禁忌搜索相结合的混合算法,利用该算法解决电力系统经济负荷分配问题。遗传算法的全局搜索能力强,但容易出现"早熟"现象,而禁忌搜索的爬山能力强,能有效避免"早熟"现象。首先利用遗传算法进行全局搜索,当算法陷入"早熟"时停止搜索,以遗传算法的结果作为初始解进行禁忌搜索,提高了初始解的质量,使禁忌搜索达到很好的效果。将该方法分别应用于某5台机组组成的发电系统和3台机组组成的发电系统进行负荷优化计算,结果与遗传算法进行比较,分析表明该算法收敛速度更快,优化成功率更高,优化结果更靠近全局最优。
A genetic-tabu search hybrid algorithm was proposed for solving power system economic load distribution（ELD）.Though having good global searching performance,genetic algorithm（GA） is prone to ＂premature＂.Tabu search（TS） is characterized by the capability of mountain climbing to avoid local optima trap.First,GA is not stopped to search in the global solution space until premature happens.The outcome of GA,which is promising solutions,is used as the initial population of TS,so TS can get good results.In order to prove the validity and effectiveness of the proposed algorithm,two test systems consisting of five generating units and three generating units were considered.The results obtained by the proposed method are compared with genetic algorithm and clearly demonstrate that the proposed method has quicker convergency speed,higher optimization success rate,and better results.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2010年第26期95-100,共6页
Proceedings of the CSEE
关键词
经济负荷分配
遗传算法
禁忌搜索
早熟
economic load distribution
genetic algorithm
tabu search
premature
