Research On Mobile Manipulator Operation Planning For Intelligent Warehousing

Traditional warehouses use manual handling that is slow and prone to errors.To address the above situation,taking the automatic sorting project of the China Robotics Competition as an opportunity,the thesis designs a mobile robotic arm to study and improve the process of its grasping of target workpieces.Through experimental analysis,the feasibility of the mobile robotic arm using AprilTag positioning to achieve accurate workpiece grasping is verified,with the following main aspects of research:(1)In response to the existing competition rules,a more realistic working task requirement is proposed and the overall design of the mobile robot arm is based on its working task.The whole machine consists of four parts: omnidirectional mobile chassis,robot arm,vision system and control system,the structure of each part is designed and suitable hardware modules are selected.(2)Kinematic analysis of the mobile robotic arm.Based on the design parameters of the robot arm a D-H parameter table is established and based on this the forward and inverse kinematic equations of the robot arm are derived.The kinematic equations were verified by means of a robot toolbox and finally its working range was found using the Monte Carlo method.(3)The trajectory planning of the motion of the robotic arm to grasp the target workpiece is studied,and the trajectory planning in Cartesian space and Joint space is introduced.Then,the trajectory planning of the whole workflow is carried out by using cubic and quintuple polynomial functions.In view of their shortcomings,the "3-5-3" mixed polynomial trajectory planning method is proposed,which combines the advantages of both and makes the simulation curve of each joint variable smoother and more continuous.For the problem of how to avoid obstacles during the robot arm’s task,RRT and its improved algorithms RRT* and reformed RRT* are investigated.After simulation analysis,the average sampling time of reformed RRT*algorithm is reduced by 66.056% and the average sampling points are reduced by 61.17%compared with RRT* algorithm.(4)The recognition and positioning principle of AprilTag is introduced,and the simulation model and environment of mobile robotic arm are built based on ROS system,and the RRT obstacle avoidance algorithm is verified.The camera is calibrated by Zhang’s calibration method,the software interface of the mobile robotic arm upper computer is designed,and the whole machine grasps the workpiece for gripping experiments.Through the experimental verification,the gripping success rate of the mobile robotic arm is 93.3%,and it runs smoothly during the working process,which is not easily affected by external environmental factors and can complete the task of gripping the target workpiece.