Research On AGV Control System Based On UWB And Visual Navigation

Automated Guided Vehicle(AGV),as the top priority of intelligent logistics in modern industrial intelligent factories,has become a hot research topic.Traditional AGVs have drawbacks such as low navigation accuracy,low robustness,poor flexibility,and harsh environmental requirements,which cannot meet the requirements of modern industry for high navigation accuracy,strong robustness,high intelligence,and high flexibility of AGVs.Therefore,in order to solve the above problems,this article conducts research on AGV control systems based on the fusion of Ultra Wide Band(UWB)and visual navigation.The main content is as follows:(1)The driving mode of AGV is determined as two wheel differential drive,the AGV coordinate system is defined and its differential drive kinematics model is established,the advantages and disadvantages of the common control modes of AGV are compared and analyzed,the fuzzy PID control is determined as the control mode of this paper according to the performance requirements,and the grid map is determined as the navigation map of this paper,and the construction process of the grid map is deduced in detail.(2)Research on UWB indoor positioning algorithm.Firstly,perform amplitude limiting filtering on bilateral ranging data to eliminate coarse errors,and then perform Kalman filtering on the data to obtain stable ranging data.It is proposed to use line intersection algorithm instead of the original trilateral algorithm to calculate two-dimensional positioning coordinates,and then optimize the target position based on median mean filtering on the calculated positioning coordinates.A UWB positioning system is built for experiments,The experimental results show that the algorithm used in this article effectively improves the accuracy of UWB indoor positioning;(3)Research was conducted on visual navigation technology and path planning algorithms,and an improved FAST algorithm was proposed to detect feature points,effectively avoiding regional clustering of feature points.The RANSAC algorithm was introduced into the rough matching process of feature points using brute force matching,eliminating erroneous information during the matching process and obtaining accurate feature point matching information.A non-linear optimization method was proposed to solve the pose of AGV,obtaining more accurate real-time pose information of AGV.In terms of path planning,this article uses the A * algorithm and DWA algorithm for global path planning and local path planning for AGV,respectively.(4)A differential AGV experimental platform based on UWB and visual navigation fusion was built,and the expected planned driving path following experiment was conducted on the AGV.The experimental results showed that the AGV designed and built in this paper has good path following performance,and also verified the feasibility of various algorithms in this paper.