| Automated guide vehicle(AGV)is the basic "foundation" of material scheduling and handling in intelligent chemical plants.Compared with the traditional fixed track AGV,the more flexible AGV with independent map building,non fixed path navigation and independent avoidance of unknown obstacles is a more preferred choice for intelligent chemical plants in the future.Starting from the actual needs,this paper designs and implements an AGV mapping and navigation system based on robot operating system(ROS)based on the requirements of intelligent chemical plant for AGV high autonomy and high flexibility.The main contents of this paper are as follows:1、The AGV without sign autonomous navigation is the design goal of this paper.Based on the system can meet the design requirements of AGV for autonomous avoidance of unknown obstacles without fixed path,the overall framework of AGV mapping and navigation system based on ROS platform is designed.The system function design is divided into two parts: mapping and navigation,and the function algorithm is designed and realized.2、The cartographer laser mapping algorithm is used to construct the navigation a priori map.Firstly,the basic principle and architecture of cartographer algorithm are introduced and analyzed.Aiming at the accuracy of odometer data,EKF is used to fuse IMU data and pre correct the odometer to improve the accuracy;Integrating the velocity data of IMU and odometer to calculate the pose increment and improve the pose estimation accuracy;For the cumulative error of the overall data and the observation noise of lidar observation,the map mosaic method is used to splice the local map with excellent observation angle with the global map.The feasibility and effectiveness of the functional algorithm are verified by the mapping experiment in the simulation environment;By comparing the composition effect of the improved algorithm before and after,it is proved that the improved algorithm improves the accuracy of map construction.3、The dynamic path planning method of A-star algorithm combined with TEB algorithm is designed,and the Adaptive Monte Carlo positioning(AMCL)method is used for navigation and positioning.The algorithm is simulated in the simulation environment to verify the feasibility of the design method and meet the requirements of system function design.4、The function algorithm design of the whole system is realized in the real environment.The real AGV mobile platform is built with ROS framework,and the mapping and autonomous navigation experiments in the real environment are carried out.The experimental results prove the superiority of the improved cartographer algorithm composition effect and the practical feasibility of the system navigation function algorithm design. |
