基于超图的自动导引小车动态作业调度问题研究

发布时间：2018-02-09 21:13

本文关键词： 超图自动导引小车动态调度路径规划时间窗　出处：《沈阳大学》2015年硕士论文　论文类型：学位论文

【摘要】：当今,自动化物流系统、柔性制造系统(Flexible Manufacturing System,FMS),自动化立体仓库等的发展十分迅速,自动导引小车(Automated Guided Vehicle system,AGVs)在生产系统中占重要的地位,作为柔性生产的重要手段,得到越来越多的应用。AGVs替代了传统的人力、拖车、传送带的输送,不存在传统车间物流运输的局限性,降低了人力成本,方便了自动化管理,提高了系统的柔性和灵活性,提高了企业效益。AGV的应用面对的是一个完整的系统。就系统而言,AGV路径规划及多AGV避碰就显得尤为重要。路径规划是AGV研究中一个重要的课题。本文结合辽宁省科技计划项目“装备制造综合自动化省级重点实验室建设”以及市场对AGV的实际需要,研究了柔性制造系统中AGV路径规划算法。结合知识和实际生产对自动导引小车动态作业调度进行研究,主要从这几个方面:(1)将AGV运行环境抽象成赋权有向图,对图论进行理论储备,且对无环赋权有向图建立路径规划模型。基于超图的路径规划模型采用遗传算法求解,并对其进行了仿真分析。(2)研究对单AGV路径规划选择了基于有向图的算法,通过分析Dijkstra算法、A*算法、极小代数的路径规划方法的特点,选择极小代数求解单AGV路径规划问题,考虑实际系统的运行状况该算法更加适合实际应用。(3)针对多AGV系统,其中有冲突和死锁问题,提出时间窗的动态路径规划算法,能有效地避免车—车冲突、碰撞等问题,并且得到了时间最优的路径。通过对时间窗的更新、时间计算,能达到多辆车一起处在工作状态,车与车之间不会碰撞。仿真表明算法有效,能处理很多情况下的多车运行问题,其具有可行性、正确性。
[Abstract]:Nowadays, the development of automatic logistics system, flexible Manufacturing system, automatic three-dimensional warehouse and so on is very rapid. Automated Guided Vehicle system (AGVs) plays an important role in the production system, as an important means of flexible production. More and more applications. AGVs replace the traditional labor, trailers, conveyor belt transportation, there is no limitation of traditional workshop logistics transportation, reduce labor costs, facilitate automatic management, improve the flexibility and flexibility of the system. The application of improved enterprise benefit. AGV is faced with a complete system. In terms of system, it is particularly important to plan the path of AGV and avoid collision with multiple AGV. Path planning is an important subject in the study of AGV. This paper combines with Liaoning Province. "Construction of Provincial key Laboratory of Integrated equipment Manufacturing Automation" and the actual needs of the market for AGV, a scientific and technological planning project, In this paper, the AGV path planning algorithm in flexible manufacturing system is studied. Combining knowledge and actual production, the dynamic job scheduling of automatic guided vehicle is studied. The AGV running environment is abstracted into weighted directed graph from these aspects. The graph theory is theoretically reserved, and the path planning model is established for the unannular weighted directed graph. The path planning model based on hypergraph is solved by genetic algorithm. The algorithm based on directed graph is selected for single AGV path planning. By analyzing the characteristics of Dijkstra algorithm and path planning method of minimal algebra, we select minimal algebra to solve the single AGV path planning problem. Considering the running condition of real system, the algorithm is more suitable for practical application. (3) for multi-#en0# systems, where there are conflicts and deadlocks, a time-window dynamic path planning algorithm is proposed, which can effectively avoid vehicle-vehicle conflicts and collisions. By updating the time window and calculating the time, we can get many cars working together, and there will be no collision between the vehicles. The simulation shows that the algorithm is effective. Can deal with many cases of multi-vehicle operation problems, its feasibility, correctness.
【学位授予单位】：沈阳大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP23

【参考文献】