人工蜂群演算法是 Karaboga 在 2005 年提出的一種基於群體智慧的仿生優化演算法,為 近年來較新的熱門最佳化問題求解的演算法之一。雖然有優異的求解能力,但仍有過早收斂 和可能陷入區域解等問題的改良研究空間。 本研究提出運用最佳訊息的移動與活化等策略,以針對標準版人工蜂群(ABC)演算法的 缺點進行改良。運用最佳訊息的移動策略能加快收斂速度,活化策略則能在陷入區域解時能 提供有效跳脫的方法。 實驗結果表明,本研究提出新的運用全域最佳訊息與活化策略改良人工蜂群演算法稱作 BPABC 演算法,在單峰、多峰等數種函數問題求解效能與效率上,均優於標準版人工蜂群 (ABC)演算法。
Artificial bee colony algorithm is invented in 2005 by the Karaboga a biological-inspired optimization algorithm based on swarm intelligence, the most popular in recent years than the best one of the new algorithm for solving problems. Although better than other algorithms for solving ability, but there is still likely to fall into premature convergence and suboptimal solutions issue. We propose the use of information of the global best solution and activation strategy to modified the original ABC algorithm to improve the drawbacks. Use of information of the global best solution strategy can speed up the convergence, the activation strategy into a regional solution is able to provide exploration capability. In this paper. We present an algorithm using information of the global best solution and activation strategy called the BPABC algorithm. To improve and enhance the original ABC algorithm in solving ability. The experimental results show that the new BPABC in the uni-modal, multi-modal benchmark functions on problem-solving performance are better than the original ABC algorithm.