Design And Implementation Of AGV Multi-layer Path Planning Method In Indoor Environment

With the advancement of intelligence and automation in various industries,AGVs have also been used more and more.At the same time,higher requirements have been placed on the capabilities of AGVs.Path planning is one of the basic capabilities of AGVs.It greatly determines the operating efficiency of the AGV system.Past research is mainly aimed at path planning in outdoor environments.Due to the large differences between indoor and outdoor scenes,many outdoor path planning methods cannot be directly applied to indoor environments.Therefore,path planning in indoor environments must be studied,combined with indoor environments.The characteristics of the environment and indoor motion can realize the path planning algorithm more suitable for the indoor environment.To this end,this paper studies AGV path planning in indoor environment.This paper first analyzes the structure of the two-wheel differential AGV,establishes its kinematic model,analyzes the influence of key parameter errors in the model,and proposes the corresponding correction method to provide support for the execution of AGV path planning.Then,starting from the characteristics of the indoor environment and the planning requirements,an indoor initial map is established using Fast SLAM based on improved particle filtering,and a layered cost grid map is realized by customizing grid map units to provide a basic map representation for path planning.Then,a hierarchical planning is proposed for the path planning task,and the global path planning and local path planning are used to jointly plan the robot path.For global planning,an improved global planning algorithm combining the A* algorithm and the RRT algorithm is proposed.This algorithm uses different algorithms under different resolution maps,and the A* algorithm under the low resolution map provides the RRT algorithm under high resolution.The biased sampling guidance domain improves the efficiency of the global path while retaining the efficiency and flexibility of the RRT algorithm.For local planning,the local map is dynamically updated by combining the positioning functions of the global grid map and odometer,and the curve similarity is used to measure the deviation of the local planned path from the global path,which ensures flexible and safe local path planning at the same time.This ensures that the two do not deviate sharply,avoids the re-planning problem caused by the failure of the global plan,and improves the efficiency of path planning.Based on the above technical research,this paper implements a hierarchical path planning system on ROS,and builds an indoor environment for experiments based on Gazebo simulation.The effectiveness of the multi-layer path planning system is verified by Gazebo physical simulation experiments.The system can build an environment map in an indoor environment,integrate the global and local planning results,and control the AGV to travel along the planned path.