基于分享学习和柯西变异的多目标人工蜂群算法

A Multi-Objective Artificial Bee Colony Algorithm Based on Sharing Learning and Cauchy Mutation

提出一种基于分享学习和柯西变异的多目标人工蜂群算法.该算法在基本人工蜂群算法中引入精英策略,即蜜蜂在更新食物源过程中,在随机选择邻居的同时,将全局最好个体以及外部档案中所有个体的平均位置作为分享学习的对象.在每次迭代结束后,对外部档案中排名位于前5%的个体进行柯西扰动,以增加解的多样性,并使得算法在求解复杂多目标优化问题时有能力跳出局部最优.在一些测试函数上的实验结果表明提出的新算法在求解多目标优化问题时,与某些经典算法相比具有一定的优越性.

An artificial bee algorithm based on sharing learning and Cauchy mutation (SLCMOABC) for solving multi-objective optimization problems is proposed in this paper. The algorithm introduces the elite strategy in the basic artificial bee colony algorithm, that is, in the process of updating food sources, not only bees randomly select neighbors, but also the global best position and the average position of all individuals in the external archive are also considered as a reference for learning. At the end of each iteration, Cauchy mutation is performed on the individuals ranked in the top 5% in the external archive to increase the diversity of the solutions and make the algorithm have the ability to jump out the local optimum when solving complex muti-objective optimization problems. The proposed algorithm proves to be competitive in dealing with multi-objective problems in comparison with other state-of-the-art algorithms for some test instances.