摘要
通过采用群体化策略和竞赛奖励制度,提出一种集群竞赛优化算法,该算法的基本思想可以归纳为竞争择优、胜者奖励、向优集群和保持多样,指出该算法与其它集群智能方法之间的联系与区别。采用多个经典测试函数对该算法进行评价并与其它优化方法进行比较。比较结果表明,平均起来,该算法优于粒子群优化算法和一种进化优化方法。
A swarming contest optimization algorithm is proposed by using colonization strategy and contest encouragement policy. The basic idea of the algorithm can be summarized as competing for better, rewarding the winner, swarming about the best and keeping diversity. The relationships and differences between the algorithm and other swarm intelligence methods are described. This algorithm is evaluated on a number of classical test functions and compared with other optimization methods. The comparisons show that, on average, this algorithm performs better than particle swarm optimization algorithm and one evolutionary optimization method.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2005年第2期142-147,共6页
Pattern Recognition and Artificial Intelligence
基金
国家863计划(No.2001AA413420)
山东省自然科学基金(No.2003G01)
关键词
进化计算
集群智能
集群竞赛优化
Evolutionary Computation
Swarm Intelligence
Swarming Contest Optimization