一种基于邻居自适应的多目标元胞遗传算法

Multi-objective cellular genetic algorithm based on adaptive neighbors

针对现有多目标元胞遗传算法存在邻居单一固定、不能适时变化的缺点,提出一种基于邻居自适应的多目标元胞遗传算法。该算法在经典多目标元胞遗传算法的基础上引入邻居自适应机制,动态调节邻居结构,使算法不断寻找全局搜索与局部寻优之间的平衡点。最后,与现有流行的其他多目标进化算法作比较,通过对不同类型的20种基准测试函数问题进行测试,证明该算法具有良好的收敛性和扩展性。

In order to solve the current problems which are that the neighbors in multi-objective cellular genetic algorithm are always fixed and can＇t be changed,this paper presented a multi-objective cellular genetic algorithm based on adaptive neighbors. The algorithm brought the adaptive strategy of neighbors into the classical multi-objective cellular genetic algorithm,so as to adjust the structures of neighbors to maintain the tradeoff between the exploration and exploitation. Finally,it compared the algorithm with the present popular multi-objective evolution algorithms and the results show that it has better convergence and expansibility on testing 20 kinds of different benchmark function problems.