Research On Omni-directional Mobile Robot Navigation System Based On Mecanum Wheels

In the factory storage environment,the material transfer has always been an important research part for improving the production efficiency of factories.With the development of robotics technology,various factories have begun to use mobile robots to carry out operations in order to increase material transfer and production efficiency and reduce production costs.The mobile robots that have been used in industrial sites now include AGVs,smart mobile robots,etc.Their work is mainly to sense the surrounding environmental information of robots by using sensors on robot to transfer material and production,along a path given or planned by the robot,which is transported to the desire location to achieve the transfer work of the factory automatic.Limited by the narrow,variable and contaminated environmental of the factory,traditional two-wheels-differential robot is not suited for the production environments in industrial.Based on the existing status of mobile robots applications in industrial field,this paper analyzes the development of mobile robots and intelligent navigation systems,and designs an omni-directional mobile robot platform based on Mecanum wheels,and builds an omni-directional mobile robot low-level-control system to realize the movement of the robots;then a hybrid resampling algorithm based on system re-sampling is applied for the robot localization algorithm;to gain the advantages of the omni-directional movement,global path planning algorithm is optimized to short the length of the path;and the Elastic Bands algorithm is used to improve the local planning algorithm to realizes the position adjustment function in the robot movement.Finally,the mobile robot platform is connected to the ROS system.The robot's control,positioning accuracy and navigation effect are tested,and the effectiveness of the improved algorithm is verified through experiments.The main contents of this article are:1.Analyzing the development of existing mobile robots,build an omni-directional mobile robot based on the Mecanum wheels,deducing the movement model of the omnidirectional mobile robot and the positive and inverse kinematics equation of the robot,building a low-level-control system,and integrating 2-lasers' data.2.Analyzing the localization algorithms commonly used in intelligent mobile robots,and using the hybrid re-sampling algorithm based on system re-sampling to improve the particle degeneration problem when re-sampling in particle filter algorithm.This improves particle diversity,speeds up the convergence of particle numbers,and improves computational efficiency.The localization algorithm has been proved to be effective through simulation and positioning accuracy experiments.3.Analyzing the global path planning algorithm commonly used in mobile robots,for A* algorithm can not obtain the shortest length of the path,the Bresenham algorithm and Bezier curve are used to optimize and improve the path which obtained by the A* algorithm.Finally,a simulation comparison experiment was conducted to verify the advantages and effects of the improved algorithm.4.To use the advantage of omni-directional mobile robot,the Dynamic Window Approach(DWA)algorithm is analyzed and used for linear velocity sampling,the Elastic Bands algorithm adjusts the pose for local path planning,and it has the characteristics of omni-direction moving and improves the motion efficiency.The simulation and experiment show that the improved algorithm can give full play to the characteristics of the omnidirectional mobile robot,and has excellent obstacle avoidance effect.