Design And Implementation Of Multi-tracked Omnidirectional Mobile Robot Based On ROS

Robots have been widely applied in production and life of human beings,without connection cables and the fixed installation positions mobile robot making the robot activities more flexible,and it can replace human to complete more work.Omnidirectional mobile robot plays an important role in the field of mobile robot,with the further development of production automation and the demand for omnidirectional mobile robots in emerging service industry,especially in the harsh ground environment and heavy load conditions,the traditional omnidirectional mobile robot can no longer meet the requirements of complex working environment.Aiming at the shortcomings and limitations of the existing omnidirectional mobile robots,a multi-tracked omnidirectional mobile robot is designed and implemented based on the ROS(Robot Operating System)platform.It can not only be used in a relatively harsh road environment or heavy load conditions,but also can realize the autonomous positioning and navigation of the robot based on laser radar under the ROS framework.Inspired by the motion principle of the traditional omnidirectional mobile robot with deflection wheel,and combining the motion principle of passive deflection,the robot uses 8 tracks to form 4 groups of tracked unit's motion structure to achieve omnidirectional mobile,further analyzing the omnidirectional direction of the robot.The structure model of omnidirectional motion of robot is analyzed,and the mechanical structure design of robot is described in detail.According to the principle of the robot,the control circuit is designed,the function of the controller and the circuit is introduced.The PID control algorithm is used to control the speed of each track of the robot.By using the compound PID control method,the deflection angle control algorithm of the caterpillar unit is designed to realize the independent control of the motion direction and speed of each track unit of the robot.The kinematics model of the mobile robot is studied to establish a kinematics model of a multi-track omnidirectional mobile robot,and a robot attitude calculation algorithm is deduced.According to the kinematics model,the algorithm of robot pose calculation is derived.It can solve the omnidirectional motion control instruction of the robot as the track speed of each crawler and the deflection angle of each track unit.The robot kinematics model is simulated by Matlab,and the trajectory of the robot and the deflection state of the track element are calculated under the omnidirectional motion control command.The simulation results verify the correctness of the kinematics model and the attitude algorithm.The attitude settling algorithm is implemented in the robot controller,and the omnidirectional motion experiment of the fixed trajectory is carried out.The trajectory estimation algorithm is used to calculate the trajectory of each track unit,and the distance variation of the caterpillar unit is analyzed and the location error of the robot is estimated during the omnidirectional movement.Experiments show that the robot has good omnidirectional mobile ability.The ROS framework is applied to the control of the robot by equipping with devices such as a Raspberry Pi 3B and a wireless router on the robot.First of all,in the Raspberry Pi deployment ROS,the robot's bottom controller connects to raspberry pie through serial port,improves the bottom controller program,releases robot status information and subscribes control command message.Secondly,ROS is deployed in a multi-level operating environment,build the URDF(Unified Robot Description Format)description file on the external computer terminal.The intermediate message upload node is designed,the coordinate transformation and odometer information of robot are released,and the running state of robot is displayed in real time in RViz.In ROS,we develop the robot control command transformation node,and realize the omnidirectional mobile remote control through the handle.After introducing the principle of laser radar,the data acquisition work of the laser radar and camera mounted on the robot was completed.In the framework of ROS,the robot SLAM(Simultaneous Localization and Mapping)and autonomous navigation problems are explored.Firstly,the types of robot environment maps are introduced and the principle of laser radar SLAM is expounded,and the SLAM experiment of the developed robot is used to create grid map of operation environment.Then,the principle of Monte Carlo localization is introduced,and combined with the static map created by SLAM to achieve the positioning of the robot in the map.In the end,the robot's global path planning and local path planning algorithm are briefly described,and the autonomous navigation and obstacle avoidance of the robot are realized based on the function package in the ROS.