Vision Guidance Of AGV And Its Path Planning Strategy Study

AGV (Automated Guided Vehicle) is a kind of material carrier with many advantages, such as high degree of automation, flexible operation, safe and reliable. It is the key equipment of flexible manufacturing systems (FMS), flexible carring system, automated warehouse and modern logistics system .In recent years, the AGV based on the vision navigation developes rapidly, because it is easy to maintain and change road, prevents electromagnetic interference and identify multistage and path branch. It has been the focus of AGV technique research.In this paper, we analyzed and studied the hardware and the key technology of the vision navigation AGV-Reinovo on the basis of the existing literatures. the main contents are summed up as follows:Firstly, we analysised the the whole structure and hardware of vision guided vehicle. And its main hardware was divided into three parts:perception directors and decision-making. On the basis of that, here is the electric diagram of AGV system.Secondly, the procedure of AGV image preprocessing was detailed and the principle of Hough transform was introduced. We improved it and applicated to the identification of AGV identifier. The experiment with VisualC++6.0 programming proved the validity of the improved Hough algorithm.Thirdly, responsed to the requirement of the designed of AGV system, we studied the design of controller and multi-AGVs path planning.This paper used the algorithm of least-squares to extract the path of AGV and angle deviation signals based on image segmentation. We used the optimal controller to realize the function of AGV walking after the mathematical model of the system was established and finished the simulation experiment. To solve the problem of NP and systematic efficiency reduce under the condition of system path node increasing, we presented the step control algorithm. Example analysis showed the effectiveness of the proposed algorithm.Finally, this paper analyzed the software structure of AGV system and emphatically introduced the function modules of the vehicle control system. On the basis of that, we programed and debuged the module of station recognition based on the improved Hough transform algorithm.