Vision Guided AGV Control System Design

In recent years, with the efficiency of the logistics system accelerating and the developing of the enterprise automation, the Automated Guided Vehicle (AGV) has been used widely already. Generally, AGV can be divided into two types according to the different guided methods, wired guidance and wireless guidance. In the latter, Vision Guided is deemed to have a good prosperous future for its simple path setting, low costs, easy maintenance, good flexibility and so on. Now, the main factors affecting AGV performance include three sides: Image processing, Motion control and Path planning. This paper will make an intensive study of them and design a system which meets a certain performance and industrial needs.This paper presents a good Image Processing method to get the path parameters, models AGV and designs the Linear Quadratic Regulator (LQR) and Fuzzy Controller separately. Based on the contrast analyses of these two controllers, we finally choose the Fuzzy Controller. In this case, AGV can track the path line stably and quickly. In addition, we also plan the running path, which can be set and managed diversified. At last, the software of AGV is finished and tested in this paper. After running test, the system designed in this paper has a good stability and real-time performance while the top speed is 0.8m/s. In the Image Processing Part, we use the Open Computer Vision Library (OpenCV).The system designed in this paper has reached our required performance index and the cost has been well controlled also, so it can be used and put into production.