基于动态规划与免疫遗传算法的多穿越巷道拣选路径规划研究  

Dynamic Programming and Immune Genetic Algorithm-based Multi-cross Aisles Order Picking Path Planning Studies

作　　者：靳萌[1] 穆希辉 杜峰坡 陈建华[1] 

机构地区：[1]中国人民解放军军械工程学院弹药工程系,石家庄050003 [2]总装备部军械技术研究所,石家庄050003

出　　处：《计算机测量与控制》2013年第11期3120-3123,共4页Computer Measurement &Control

摘　　要：针对多穿越巷道布局立体仓库的拣选路径规划问题,总结了现有的成熟路径生成策略,分析了现有路径生成规划方法的优劣,在此基础上提出了一种结合动态规划方法的启发式搜索算法生成拣选路径;该方法调和了动态规划空间开销大与启发式搜索时间开销大的矛盾;利用限制迭代次数动态规划生成初始种群,免疫算子与遗传算子寻优求解,有效解决了多穿越巷道布局下的拣选路径规划问题;通过数值仿真实验验证,算法有效节约了计算机空间资源,并在可接受的时间内生成了可接受的路径规划结果;虽然算法还存在很多不足,但为分布式布置信息化立体仓库的车载有限资源计算机实时生成路径提供了一种可行的思路。For stereoscopic warehouse with multi cross aisles layout order picking path planning problem, summarizes the existing mature path generation strategy, analysis of the pros and cons of the existing path generation planning method, Based on this, advances a kind of combining with dynamic planning method of heuristic search algorithm to generate picking route. This method is to reconcile the dynamic planning space overhead and heuristic search time costs big contradiction. Using the limited number of iterative dynamic programming to gen erate the initial population, the immune operator and genetic operators optimization solution, and effectively solves the order picking path planning problem under multiple cross aisles layout. Through numerical simulation experiment, the algorithm saves computer space resources effectively, and at an acceptable time endogenous became an acceptable path planning result. Although algorithm also has many shortcom ings, but limited resources for distributed information of stereoscopic warehouse vehicle computer real time generate path provides a feasible idea.

关 键 词：多穿越巷道 立体仓库 动态规划 免疫遗传算法 

分 类 号：TP271[自动化与计算机技术—检测技术与自动化装置]