基于有限缓存区的多任务并行分拣调度研究

通过对某物流车间的实际调研,将自动化立体仓库出货台空间限制作为优化问题的约束条件,建立订单并行分拣模式下堆垛机调度问题的模型,并采用蚁群算法进行求解。在求解过程中,根据问题假设设定了算法相关的状态转移概率公式,并采用动态更新信息素浓度的改进型方式避免传统蚁群算法早熟的情况。最后根据工厂的实际订单信息给出了算例,并通过两种不同算法和不同参数设置的比较,说明通过蚁群算法求解该优化问题的有效性。数值试验显示该蚁群算法相比传统优化算法效率提升了10.5%。

Study of Scheduling of Parallel Multi-task Picking Based on Limited Working Area

Through the actual investigation of a logistics workshop,it established the stacking machine scheduling model under the order parallel picking mode,and solved this by using ant colony algorithm,with the limit of shipping space of AS/RS as the constraints of the optimization problem.During the solving process,it set the state transition probability formula and adopted the method of dynamic and updating pheromone concentration to avoid the premature problem of traditional ant colony algorithm.Finally,an example according to the concrete factory order was given in this thesis to demonstrate the effectiveness of ant colony algorithm through the comparison of two different algorithms and different parameter settings.The numerical experiments showed the ant colony algorithm compared to the traditional optimization algorithm enhanced 10.5% efficiency.