以群中心策略改良人工蜂群演算法

Enhanced Artificial Bee Colony Algorithm with Centroid Strategy

李元傑(Yuan-Chieh Lee)

指導教授 ： 李維平

中原大學/商學院/資訊管理研究所/碩士(2011年)

https://doi.org/10.6840/cycu201100478

摘要

人工蜂群演算法 (Artificial Bee Colony, ABC) 是於2005年由Karaboga所提出之最佳化演算法，它具有非常良好的穩定性與求解能力，控制參數少、計算簡潔而易於實現。人工蜂群演算法的效能也已經由多個實驗研究，得到了眾多學者們的肯定，並逐漸受到重視，而開始廣泛應用於各種最佳化問題，因此，近年來學者們也開始熱衷於改良人工蜂群演算法，使其成為近年來熱門的研究對象之一。
人工蜂群演算法雖然具有非常良好的穩定性，但是卻依然存在著後期過早收斂、開發精度不佳等問題。本研究在參考了眾多學者之相關改良文獻後，發現以群中心為基礎的改良方法，其發展還尚不成熟，因此，本研究提出一種新式的群中心改良策略，嘗試加強人工蜂群演算法之開發能力。
本研究共尋找了22個測試函數來進行實驗，從實驗結果得知，本研究提出之群中心策略，確實能夠有效地加強人工蜂群演算法的開發能力，使其在後期可以持續開發而不會過早收斂，在大部分測試函數上的表現都具有明顯的改善，除了能夠獲得更好的結果之外，還能夠加強其穩定性。另外，本研究也將群中心策略應用於其他學者提出之已改良人工蜂群演算法，確實也能夠加強其效能，並能夠獲得更好的結果。

關鍵字

最佳化演算法 ； 相對位置 ； 演化式計算 ； 群中心 ； 人工蜂群演算法

並列摘要

Artificial Bee Colony (ABC) was proposed in 2005 by the Karaboga, it has a very good stability and solves capability, the control parameter is few, computing simple and easy to implement. ABC is also already studied by several experiments, many scholars have been affirmed, and gradually be taken seriously, and have been widely used in various optimization problems. In recent years, scholars also start falling overboard for improvement ABC, making it become a popular object of study in recent years.

ABC, while having very good stability, but there are still premature convergence, development issues such as poor accuracy. In this study, a number of scholars in reference to the literature related improvements, we found the centroid of swarm-based improvement methods, it development is still not mature. Therefore, this study proposed a new centroid of swarm-based improvement methods to try to strengthen the development capability of ABC.

In this study, a total of 22 test functions looking to experiment, learn from the results of this study proposed improvement methods, can indeed effective in enhancing the development capability of ABC, make it can sustainable development without premature convergence, in performance on most test functions have improved significantly. in addition to able to obtain better results, but also to enhance its stability. In this study, we also applied our improvement methods to improved ABC of Other scholars proposed, can really also strengthen its capability, and can acquire better result.

並列關鍵字

Artificial Bee Colony (ABC) ； Centroid ； Evolutionary Computation ； Optimization Algorithm ； Opposition

參考文獻

[2] 林豐澤，“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，第3卷，第1期，第1-28頁，2005年。

[3] 郭信川、官佳慶，”隨機搜尋法於多極值最佳化問題之應用”，中國造船暨輪機工程師學刊，第19卷，第4期，第33-40頁，2000年。

[11] H.A. Abbass, “MBO: marriage in honey bees optimization - a Haplometrosis polygynous swarming approach,” 2001. Proceedings of the 2001 Congress on Evolutionary Computation, Vol. 1, pp. 207-214, May, 2001.

[12] B. Akay, D. Karaboga, “Artificial bee colony algorithm for large-scale problems and engineering design optimization,” Journal of Intelligent Manufacturing, 2010.

[13] B. Akay, D. Karaboga, “A modified Artificial Bee Colony algorithm for real-parameter optimization,” Information Sciences, 2010.