Research Of AGV Control System Based On Vision

The automatic guided vehicle is widely used in modern industrial production and logistics transportation.In the traditional guidance mode,the magnetic guidance mode is the main method of AGV.The deviation information is single and the control effect is worse,which makes the operation of AGV easy to produce the phenomenon of left and right wobble,then it may derails.In the actual factory workshop,the traditional PID controller's control technology is difficult to meet the complex trajectory tracking requirements of AGV.Therefore,the research on path identification and trajectory tracking control of AGV is of great value.The main research contents are made up of the following parts:1)The mathematical model of the AGV car is built and the six wheeled stable structure is adopted to the AGV.The two wheel differential drive is used as the driving mode and the driving mode is analyzed by the kinematics principle.The mathematical relationship between the distance error and the angle error and the two wheel drive mode is obtained.Therefore,the whole system equation of the required AGV is obtained.2)The machine vision guided method is used to guide the extraction of information images and the OpenCV software package is used for image processing.It includes the filtering of the original figure,the detection of the edge line,the recognition of the ribbon and the fitting of the center line of the ribbon.3)A fuzzy controller is designed based on fuzzy control theory for AGV trajectory tracking control system.Then,the traditional PID controller and the designed fuzzy controller are simulated individually by Matlab software.Comparing the process convergence curve and smoothness between the two methods.The results of simulation show that the AGV trajectory tracker based on fuzzy control can effectively solve the problem of suboptimal control under the complex operation.4)An AGV control platform based on vision is built.The AGV vehicle test was carried out in laboratory and practical application to verify the performance of AGV in visual image processing and trajectory correction.