結合群中心策略與對稱學習改良人工蜂群演算法

A Novel Artificial Bee Colony Algorithm with Centroid Strategy and Opposition-Based Learning

吳易任(Yi-Ren Wu)

指導教授 ： 李維平

中原大學/商學院/資訊管理研究所/碩士(2013年)

https://doi.org/10.6840/cycu201301012

摘要

人工蜂群演算法(Artificial Bee Colony, ABC) 是一參考蜂群採蜜的分工模式所提出的群體智慧演算法。由於ABC與其它演算法相比，有極為突出的表現，特別是在函數優化的領域顯現出優秀的性能，所以發表以來，持續受到學者關注。
人工蜂群演算法具有計算簡潔、易於實現、設置參數少等特點，同時卻也有區域搜索無法持續精進、可能陷入區域最佳解、偵察蜂探索效力不高等問題。針對ABC所存在的問題，本研究提出以群體演化中心作對稱的方式改良人工蜂群演算法，此一機制的構想是在每一迭代分散蜂群搜索的資源引領至陌生的區域，使蜂群的探索有明確的目標，能根據當前迭代的演化趨勢調整探索方針，提升演算法的收斂速度以及增加跳脫區域最佳解的機會。
實驗部分採用五種公認的測試函數，結果顯示本研究所提之方法不論在高低維度皆有不錯的成效，且在多峰函數測試時，擁有快速且良好的求解能力。

關鍵字

對稱學習 ； 群中心策略 ； 人工蜂群演算法

並列摘要

Artificial Bee Colony algorithm (ABC) is a collective intelligence algorithm which refers to bee swarm`s division of labor mode. Because ABC has more excellent performance than other algorithms especially the function optimization field, scholars have pay attention to it continuously since the first published.
The distinctive features of ABC include simple calculation, easy to implement, and low demand parameter settings. However, there are still some defects in it, such as regional search which was unable to improve continuously, falling into local optimal solutions, and low effectiveness exploring of scout bee. In this study, we adopt the opposition of group center to lead search for strange area. This strategy is beneficial to adjust search resource of algorithm. It regulates exploring principle according to current trend of evolution, promoting convergence speed and increasing opportunity for departing from local optimal solutions.
In this research, we use 5 generally acknowledged benchmark functions to test our method. The outcome demonstrates that our algorithm has good effect regardless of high and low dimension. It owns speedy and better resolving ability at multimodal functions.

並列關鍵字

Opposition-Based Learning ； Centroid Strategy ； Artificial Bee Colony

參考文獻

林豐澤（2005）。演化式計算上篇：演化式演算法的三種理論模式。智慧科技與應用統計學報，3(1)，1-28。

Alatas, B. (2010). Chaotic bee colony algorithms for global numerical optimization. Expert Systems with Applications, 37(8), 5682-5687.

Banharnsakun, A., Achalakul, T., & Sirinaovakul, B. (2011). The best-so-far selection in artificial bee colony algorithm. Applied Soft Computing, 11(2), 2888-2901.

Chen, S. M., Sarosh, A., & Dong, Y. F. (2012). Simulated annealing based artificial bee colony algorithm for global numerical optimization. Applied Mathematics and Computation, 219(8), 3575-3589.

Gao, W. F., Liu, S. Y., & Huang, L. L. (2012). A global best artificial bee colony algorithm for global optimization. Journal of Computational and Applied Mathematics, 236(11), 2741-2753.