Abstract: Fixed storage rack system is an important part of the automated warehouse.The efficiency of warehouse depends on the control strategies of order picking.According to the order picking procedure in the single aisle rack，a mathematic model is constructed with multiple constraints and the optimization aim is to minimum the storage and retrieval time.It adopts dynamic change on algorithm parameters and awaiting nodes set based on the basic ant colony algorithm，and designs an improved ant colony algorithm for the order picking path planning problem.Simulation results demonstrate the improved algorithm can solve medium or large scale problem，it also has better overall search ability and quickly astringency.

摘要： 固定货架系统是自动化立体仓库的重要组成部分，仓库运行效率主要取决于对货物进行拣选的控制策略。针对单巷道固定货架系统拣选作业过程，以存取时间最小为目标，构建了含多个约束条件的拣选作业路径规划问题的数学模型。在基本蚁群算法基础上，采取自适应调整算法参数、候选节点集合等改进措施，设计了一种改进的蚁群算法对问题进行求解。仿真实验表明该算法能够很好地解决中大规模的拣选作业问题，全局寻优能力强，收敛速度快。