| As the fast development of industrial automation, automated guided vehicle system(AGVS) is commonly used in facilities such as warehouses, distribution centers, manufacturing plants and assembly plants. In the step of system design and planning, the guide-path design is one of the key issues. In addition, vehicle routing and traffic management are very important in stage of system execution.In this paper, after detailed study on the advantages and disadvantages of unidirectional guide-path network and bidirectional guide- path network, an AGVS based on mixed uni/bidirectional guide-path network is proposed. The new kind of AGVS not only has a high efficiency brought by bidirectional guide-paths, but also makes the complex traffic management strategies not necessary. To design the mixed uni/bidirectional guide-path network, an improved genetic algorithm is implemented. The improved GA not only performs well in global optimization but also has a great performance in local optimization because it combines GA with neighborhood search algorithm. Besides, to simplify the traffic management strategy, this paper puts forward an innovate strategy which separates the bidirectional guide paths apart by using the unibidirectional flow paths. Besides, vehicle routing and deadlock prevention problems are studied in this paper. Firstly, we construct an electronic map for AGVS based on mixed uni/bidirectional flow path network. And then an improved two-staged routing algorithm is proposed to route vehicles in AGVS. Moreover, this paper presents a traffic management strategy which uses the traffic semaphore to prevent deadlock. With the strategies above, AGVs can run smoothly with no deadlock and high efficiency.Finally, approach proposed above is implemented in an actrual AGVS. We use plant simulation software to test the efficiency and feasibility for the AGVS with mixed uni/bidirectional guide-path network. Besides, the master controller for AGVS is constructed in this paper. |
PDF Full Text Request