【摘要】 为了将最优化方法与思想融入群智能优化的研究,提出了允许重复的存档和种群重置策略、聚集算子和分散算子,并分析了这些操作对算法性能的影响。在3种操作基础上,提出一种新的启发式算法——聚散优化算法(GAD)。将所提算法与4个经典启发式算法(SPSO2011,CoDE,SaDE和IGHS)相比较,用于求解20个经典的函数最优化问题。结果显示,聚散优化算法能得到较其他算法更优秀的解,从而说明了所提聚散优化算法及其操作算子的有效性。更多还原

【Abstract】 In order to integrate the optimization methods and ideas into the research of swarm intelligence, the strategies of allowing duplicate archiving and population reset, gathering operator and dispersion operator were proposed. The mathematical principles and the performance of these operators were analyzed. Based on these operations, a new heuristic algorithm, Gathering and Dispersion(GAD) optimization algorithm was proposed. Then, the proposed GAD algorithm was compared with four classical heuristic algorithmed(SPSO2011, CoDE, SaDE and IGHS)based on 20 classic benchmark. Results showed that GAD algorithm outperformed other competitors. It indicated that the proposed operations and the idea of popularizing GAD algorithm was effective.更多还原