Research On Trajectory Planning Of Wheeled Mobile Manipulator Based On Ros

The wheeled mobile manipulator is featured with both mobility and operability,the working environment is complex,moreover,high speed and high precision are frequently required when performing tasks,consequently,not only should the manipulator avoid the obstacles in the working space,but also the motion stability should be ensured.In this paper,the kinematics modeling of the proposed wheeled mobile manipulator is carried out in the first place.Secondly,the experimental platform of the wheeled mobile manipulator is established based on ROS,and after that,the trajectory planning theory research and simulation are carried out.Finally,the obstacle avoidance planning and the gripping performance are tested on the manipulator.Specific researches are as follows:(1)Kinematics modeling of the wheeled mobile manipulator.The kinematics model of the wheeled mobile manipulator is established.The transformation of the adjacent link coordinate system is analyzed by D-H parametric method.The equation of the forward kinematics and inverse kinematics of the manipulator is derived.According to the forward kinematics solution,the working space of the manipulator is obtained by using Monte Carlo method.(2)Establishment of ROS platform of the wheeled mobile manipulator.The URDF file of the robot is compiled.The motion configuration of the robot is carried out by MoveIt! setup assistant.The interface between the upper layer planning of MoveIt! and the bottom ArbotiX joint controller is designed,after which the communication between the manipulator and MoveIt! is established.The forward and inverse kinematics programs of the manipulator are compiled,and motion simulation is carried out in RViz.(3)Research on trajectory planning of the manipulator.In the joint space,the trigonometric function interpolation is used,and the parameters of the interpolation function are designed to optimize the joint trajectory,which overcomes the insufficiency of traditional double S-type velocity interpolation algorithm.In the Cartesian space,the linear and circular trajectory interpolation models of the manipulator are established,the linear and circular planning nodes are designed and simulated in RViz,smooth joint displacement and velocity curve are obtained.(4)Simulation and experiment analysis.The collision-free path planning of the manipulator is carried out by using Rapidly-exploring Random Tree algorithm.The planning environment is configured according to the actual conditions.The motion planning nodes are designed,and after that,simulation and experiment are carried out in ROS.In order to test the grabbing performance of the manipulator,the experiment of grabbing a pen is also carried out based on arbotix_gui.The research results obtained in this paper are as follows:(1)A piecewise interpolation method of trigonometric function in joint space is put forward.The parameters of the interpolation function are skillfully designed,which effectively solves the discontinuous problem of the jerk of traditional double S-type velocity interpolation function,simultaneously,the smoothness of the acceleration has improved dramatically,making the motion of the manipulator more stable.(2)The obstacle avoidance trajectory planning of the manipulator is realized based on the RRT algorithm.The experimental results show that when there is an obstacle in the workspace,the manipulator can avoid the obstacle and successfully plan a safe path from the initial pose to the target pose,which verifies the effectiveness of the RRT algorithm.