旋转货架拣选作业优化问题的新型混合遗传算法求解  

NEW HYBRID GENETIC ALGORITHM SOLVING THE ORDER-PICKING OPTIMIZATION PROBLEM OF A MULTI-CAROUSEL SYSTEM

作　　者：张攀[1] 田国会[1] 贾磊[1] 李晓磊[1] 路飞[1] 

机构地区：[1]山东大学控制科学与工程学院,济南250061

出　　处：《机械工程学报》2004年第6期34-38,共5页Journal of Mechanical Engineering

基　　金：国家自然科学基金(60104009);山东省自然科学基金(Z2000G01)

摘　　要：给出了单拣选台分层水平旋转货架系统的数学模型,分析了单拣选台分层水平旋转货架拣选作业路径优化问题(MCS-OOP)的特点。在单层旋转货架中待拣选货物的最优拣选顺序将依次出现在对整个作业中所有待拣货物的最优拣选顺序中,针对该特点,提出了层序邻域的概念及其快速局部搜索算法,同时将其与遗传算法相结合设计了一种用于解决MCS-OOP的新型混合遗传算法。仿真结果说明了该算法能够快速、稳定的求取单拣选台MCS-OOP问题的最优解,充分满足了中大规模作业要求。The mathematical model of the single pick-platform & multi-carousel system is given, and the order-picking optimization problem of it is analyzed. According to the character that the optimal picking-order of the materials in the single carousel will come forth of the optimal order of the whole assignment, a new layer-based neighborhood and its local search algorithm are presented. Then combined them with genetic algorithm, a new hybrid genetic algorithm solving the single pick-platform MCS-OOP is designed. Finally, a case of study in MCS-OOP using the new algorithm is described and analyzed. Experimental results verify that the optimum solutions are quickly and steadily achieved and the requirements of work are fully satisfied.

关 键 词：自动化仓库 旋转货架系统 混合遗传算法 层序邻域 局部搜索 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]