基于Lévy飞行的人工蜂群算法

An Artificial Bee Colony Algorithm Based on Lévy Flight

针对人工蜂群算法在后期容易陷入局部最优,且开发能力不足的缺陷,提出了一种基于Lévy飞行的人工蜂群算法,在寻找蜜源过程中,引领蜂转变成侦查蜂时,引入Lévy分布函数来调配原有的用0-1的随机数参与的贪婪算法,有效增加了算法的合理性。最后用5个有效的测试函数对改进的算法和原算法进行比较,结果表明,基于Lévy分布的人工蜂群算法较原算法在收敛速度和优化速度上都有显著提高。

The artificial bee colony algorithm in the later is easy to fall into local optimum, and the lack of ability to develop This paper proposed a kind of artificial bee colony algorithm based on Lévy flight, in the process of finding nectar, leading bees turn into scout bees, the introduction of Lévy distribution function to allocate raw some 0-1 random number in the greedy algorithm, which effectively increased the rationality of the algorithm. Finally, the improved algorithm and the original algorithm are compared with 5 effective test functions, and the results show that the algorithm based on Lévy distribution has significantly improved the convergence speed and the optimization speed.