基于时间窗的自动导引车系统路径优化研究

发布时间：2018-01-05 07:19

本文关键词：基于时间窗的自动导引车系统路径优化研究　出处：《兰州交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：AGV(Automated Guided Vehicle,自动导引车)是一种用于FMS(Flexible Manufacture System,柔性制造系统)和SW(Stereoscopic Warehouse,立体仓库)实现物料运输的轮式移动机器人。多AGV系统作为LS(Logistics System,物流系统)和FMS中的关键子系统,由于其高柔性且便于管理,被越来越多应用于物料输送。路径优化是多AGV系统应用中关键技术之一,既是研究热点又是难点,因而对其研究有重要理论和实践意义。本文在国内外学者研究基础上,对多AGV系统路径优化及协调避碰进行研究,主要研究工作包括如下:首先,以固定路线导引AGV为研究对象,确立AGV用于物料输送的工作环境,采用拓扑结构法在AGV运行环境信息已知条件下完成地图建模。搜索单台AGV最优路径时将固定时间常数τ附加于行驶路径转弯点作为转弯代价惩罚,转弯越多代价越大,并以最短行驶时间为评价指标,通过采用Dijkstra算法离线搜索单台AGV前K条最短路径。其次,将K条路径作为初始路径,提出以组合叠加方式分别将K条路径与已运行AGV行驶路径进行理想时间窗排布、组合叠加更新时间窗和冲突检测。针对不同类型制定不同冲突解决策略,并以各台AGV行驶时间总和最小为优化指标,实现离线状态下基于时间窗的多AGV无碰路径优化。同时在此基础上采用在线检测方法针对AGV行驶时的突发状况进行检测,并通过采取改变AGV通过节点优先级策略和等待策略实现多AGV系统在线协调。最后,在理论研究基础上采用Visual Studio2010软件和Visual C++编程语言建立多AGV地面控制系统仿真平台,分别对单台AGV最优路径搜索和多AGV无碰路径优化进行仿真实验。实验结果证明该算法对多AGV无碰路径优化具有有效性和可靠性,可提高系统柔性和效率,为实际生产中多AGV系统应用提供理论依据和工作指导。
[Abstract]:AGV(Automated Guided Vehicle. Auto-guided vehicle is used for FMS(Flexible Manufacture System. FMS) and SW(Stereoscopic Warehouse. Multi AGV system is the key subsystem of LS(Logistics system (logistics system) and FMS. Because of its high flexibility and easy management, it is more and more used in material transportation. Path optimization is one of the key technologies in the application of multi-#en0# system, which is not only a hot topic but also a difficult point. Therefore, it has important theoretical and practical significance. Based on the research of domestic and foreign scholars, this paper studies the path optimization and coordinated collision avoidance of multi-#en0# system. The main research work includes the following: first. Taking the fixed route guiding AGV as the research object, the working environment of AGV for material transportation is established. The topology method is used to complete the map modeling under the condition that the AGV running environment information is known. When searching for a single AGV optimal path, the fixed time constant 蟿 is appended to the turning point of the driving path as the turning cost penalty. The higher the cost of turning, and the shortest driving time as the evaluation index, using the Dijkstra algorithm offline search single AGV before K shortest path. Secondly, K path as the initial path. In this paper, the ideal time windows of K paths and running AGV paths are arranged by combination superposition, and the time windows and conflict detection are combined. Different conflict resolution strategies are developed for different types. The minimum total driving time of each AGV is taken as the optimization index. The off-line multi-#en0# collision path optimization based on time window is realized, and the on-line detection method is used to detect the burst of AGV driving. And by changing the AGV through the node priority policy and wait policy to achieve multi-#en1# system online coordination. Finally. On the basis of theoretical research, the simulation platform of multi-#en3# ground control system is established by using Visual Studio2010 software and Visual C programming language. The simulation experiments of single AGV optimal path search and multi-#en1# no-collision path optimization show that the algorithm is effective and reliable for multi-#en2# no-collision path optimization. It can improve the flexibility and efficiency of the system, and provide theoretical basis and work instruction for the application of multi-#en0# system in actual production.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP242

【相似文献】