第2类U型装配线平衡问题的双阶段蚁群算法

摘要

针对电子、汽车等行业中普遍存在的第2类U型装配线平衡问题(UALBP-2),提出了一种双阶段蚁群算法.强调全局搜索的第一阶段算法利用探路蚁,根据操作选择和分配策略以及迭代压缩机制快速得到问题的较优解,减小搜索空间;注重局部搜索的第二阶段算法利用搜索蚁,根据所提的信息素减小更新策略在包含最优解且不断减小的搜索空间中搜索各工位的不同精英负载,基于精英复制策略利用精英蚁将其组合为问题的可行解.对18个标杆算例的33个实例的求解结果验证了所提算法的有效性和稳定性.%A two stage ant colony optimization for the type 2 of U-shaped assembly line balancing problem (UALBP-2) was proposed,which is widespread in the electronics and automobile industry.In the first stage algorithm with the high capability of global search,a better feasible solution is obtained by the scout ants according to the task selection strategy,the task assignment strategy and the iteration compress mechanism.The search space is decreased according to the solution.In the second stage algorithm with the high capability of local search,different elite station loads are searched by the pathfinding ants according to the update strategy of decreasing pheromones.The elite station loads of every station are grouped together into the feasible solutions of UALBP-2 by the elite ants according to the elite copy strategy.The computational results of 33 instances from 18 benchmark examples verify the effectiveness and the stability of the proposed algorithm.