Abstract: At present,stereoscopic warehouse is widely used in textile workshops.Aiming at the low scheduling efficiency of a popular and general multi-task stereoscopic warehouse in textile enterprises,a set of warehouse scheduling methods was proposed from three aspects:task advance scheduling,storage location allocation and task sequence optimization.To solve the problem of unclear working time of different workers in manual operation area,a method for predicting working time of workers based on the autoregressive moving average model was proposed.In view of the unreasonable allocation of storage location,a multi-objective allocation model was established,and a solution method based on simulated annealing algorithm was proposed.For the low execution efficiency caused by the accumulation of task sequence,an improved genetic-particle swarm optimization algorithm for task sequence reordering was presented.The experimental comparative analysis showed that the proposed method could effectively improve the scheduling efficiency of the stereoscopic warehouse.

Key words: stereoscopic warehouses, time prediction, storage location allocation, task sequence optimization, texile workshops

摘要： 当前立体仓库正大规模应用于纺织车间。针对一种新型通用的纺织企业多任务类型的立体仓库调度效率低的问题,从任务提前调度、货位分配以及任务序列优化3方面提出一套立体仓库调度方法。首先,针对工人操作时间不明确的问题,提出一种基于自回归滑动平均模型的工人操作时间预测方法;其次,针对货位分配不合理的问题,建立多目标的货位分配模型,并提出一种基于模拟退火算法的货位分配求解方法。最后,针对任务序列累积导致执行效率低的问题,提出一种改进的遗传-粒子群算法的任务序列优化方法。实验对比分析表明,所提方法可有效提升立体仓库调度效率。

关键词: 立体仓库, 时间预测, 货位分配, 任务序列优化, 纺织车间