Research On Autonomous Navigation Technology Of AGV Multi-Information Fusion In Printing Wisdom Factory

With the development of the national economy and the continuous improvement of people's living standards,the packaging and printing industry serving all sectors of society is playing an increasingly important role in the national economy and has become a supporting industry that has an important influence on economic and social development.However,my country's packaging and printing companies still use labor-intensive production models,facing problems such as low management levels,high labor costs,and serious waste of raw materials.It is imperative to realize intelligent upgrades.AGV is the "bridge" for the intelligent scheduling of materials in the intelligent packaging and printing factory,and precise positioning and navigation are the prerequisite and guarantee for its application in the intelligent packaging and printing factory.Therefore,according to the requirements of smart packaging and printing factories for AGV's high autonomy and flexibility,this article studies the AGV multi-information fusion autonomous navigation technology.The main research contents are as follows:(1)The two-dimensional lidar real-time positioning and mapping algorithm combining odometer and IMU information is studied,and the high-precision real-time positioning of AGV is realized.ROS-melodic was used to build a physical simulation environment for a smart packaging and printing factory,and a two-dimensional grid map was constructed using Gmapping and Cartographer mapping methods,respectively;Aiming at the problem of inaccurate mapping caused by the accumulated error of the odometer,the odometer and inertial navigation are used for information fusion,and the fused information and lidar information are used in Gmapping for mapping,which improves the accuracy of the map.(2)Designed a combined path planning algorithm based on improved A*algorithm and DWA algorithm.The four search directions of the A*algorithm are extended to eight search directions,and a global path planning method based on the improved A*algorithm is designed.Simulation experiments show that the improved algorithm has a shorter search path;Aiming at the problem of AGV real-time obstacle avoidance,The local path planning method based on Dynamic Window Approach(DWA)is studied.The simulation experiment results show that the DWA algorithm can effectively avoid obstacles in time.The simulation experiment results show that the DWA algorithm can effectively perform real-time obstacle avoidance;By combining two path planning algorithms,a combined path planning algorithm for AGV in smart packaging and printing factories is designed.(3)Researched the solution of single AGV multi-task assignment and multi-AGV autonomous navigation.Aiming at the problem of single AGV multi-task allocation in smart packaging and printing factories,the ant colony algorithm is used to optimally solve the problem;Aiming at the problem of autonomous navigation of multiple AGVs in smart packaging and printing factories,the "leader-follow" method is used to control the three AGVs in a line-shaped formation.Through the autonomous navigation of the leader AGV,follow the movement to realize the multi-AGV Autonomous navigation.