摘要
针对最大-最小蚂蚁系统在求解顺序流水车间调度问题时易陷入局部最优的问题,对提升算法求解质量进行了研究。将最好-最坏蚂蚁系统的正负反馈机制、变异机制和模拟退火算法中Metropolis准则引入到最大-最小蚂蚁系统的信息素更新环节中,并引入局部寻优,提出了一种混合蚁群算法。在典型测试集上的实验表明,新算法在顺序流水车间调度问题上所得结果普遍优于最大-最小蚂蚁系统,具有更好的全局寻优能力。
Aiming at the requirement that Max -Min Ant System（ MMAS）plunges into local optimum easily when applied to permutation flowshop scheduling problem, it introduces the positive and negative feedback, mutation rules suggested by Best - Worst Ant System （BWAS） and metropolis rule suggested by Simulated Annealing （SA）, provides the solution quality of MMAS and proposes a hybrid ant colony system. The experiments show that the results of new algorithm are widely better than MMAS. New algorithm has better global search ability.
基金
国家自然科学基金资助项目(71171114)
