论文部分内容阅读
蚁群算法是一种模拟进化算法,具有很强的全局搜索能力.本文提出一种自适应的并行蚁群算法(A-PACO),该算法可以根据不同的搜索阶段,自适应确定参数的最优组合,在一定程度上避免停滞现象的出现并加速算法收敛.而且自适应的迁移策略可以较大丰富系统多样性的同时也较大降低子蚁群间的通信量,有效提高算法的搜索质量和缩短算法的运行时间.最后选用中国 CHN144问题对该算法进行检验,结果显示该算法具有较好的稳定性和较快的收敛速度.
Ant colony algorithm is a simulated evolutionary algorithm with strong global search ability.This paper proposes an adaptive parallel ant colony algorithm (A-PACO), which can adaptively determine the parameters of To some extent, avoids the appearance of stagnation and accelerates the convergence of the algorithm.And adaptive migration strategy can greatly enrich the diversity of the system and also greatly reduce the communication between the sub-ant groups, and effectively improve the search quality of the algorithm And shorten the running time of the algorithm.Finally, the algorithm is tested by the Chinese CHN144 problem, the result shows that the algorithm has better stability and faster convergence speed.