基于正交实验设计的人工蜂群算法

人工蜂群算法是近年来提出的较为新颖的全局优化算法,已成功地应用于解决不同类型的实际优化问题.然而在该算法及相关的改进算法中,侦察蜂通常采用随机初始化的方法来生成新食物源.虽然这种方法较为简单,但易造成侦察蜂搜索经验的丢失.从算法搜索过程的内在机制出发,提出采用正交实验设计的方式来生成新的食物源,使得侦察蜂能够同时保存被放弃的食物源和全局最优解在不同维度上的有益信息,提高算法的搜索效率.在16个典型的测试函数上进行了一系列实验验证,实验结果表明:1) 该方法能够在基本不增加算法运行时间的情况下,显著地提高人工蜂群算法的求解精度和收敛速度;2) 与3种典型的变异方法相比,有更好的整体性能;3) 可作为提高其他改进人工蜂群算法性能的通用框架,具备有良好的普适性.

Artificial Bee Colony Algorithm Based on Orthogonal Experimental Design

Developed in recent years, artificial bee colony (ABC) algorithm is a relatively new global optimization algorithm that has been successfully used to solve various real-world optimization problems. However, in the algorithm, including its improved versions, the scout bee usually employs the random initialization method to generate a new food source. Although this method is relatively straightforward, it tends to result in the loss of the scout bee's search experience. Based on the intrinsic mechanism of ABC's search process, this paper proposes a new scheme that employs the orthogonal experimental design (OED) to generate a new food source for the scout bee so that the scout bee can preserve useful information of the abandoned food source and the global optimal solution in different dimensions simultaneously, and therefore enhancing the search efficiency of ABC. A series of experiments on the 16 well-known benchmark functions has been conducted with the experimental results showing the following advantages of the presented approach: 1) it can significantly improve the solution accuracy and convergence speed of ABC almost without increasing the running time; 2) it has better performance than other three typical mutation methods; and 3) it can be used as a general framework to enhance the performance of other improved ABCs with good applicability.