Research On Motion Planning Of Mobile Manipulators In Indoor Environment

Currently,the development of service robot and artificial intelligence has become the focus of research all over the world,especially the study of mobile manipulator with wheeled chassis is a hot issue.On one hand,there is a broad prospect of application scenarios and actual requirements.On the other hand,the coupling between the manipulator and the mobile chassis also brings some challenges to the precise control of the mobile robot.In this thesis,the author built a mobile manipulator based on the RIA-E100 mobile platform and the Cobotta 6-DOF manipulator,studied the movements of the mobile platform and manipulator,and proposed an improved planning algorithm for the mobile platform and the manipulator respectively,aiming at the actual requirements of the library application scenarios,which can realize many functions,such as autonomous positioning,map building,navigation,obstacle avoidance and planning,etc.It can be used as a library service robot to relieve the pressure of the staff.The main research contents of this thesis include the following aspects:1.System design of mobile manipulator.Firstly,the hardware selection of mobile platform,manipulator and sensor was analyzed,completing the overall structure design.Secondly,the kinematic analysis was carried out to study the driving process of the two differential wheels of the mobile platform.Based on the basic coordinate system of the mobile platform,the improved D-H parameter method was selected to complete the modeling of the manipulator.Combined with Pieper's rule,the geometric and analytical method was used to simplify the solution of inverse kinematics.2.Path planning of mobile platform and an improved J-A* algorithm of the global path planning.Firstly,taking the laboratory as the experimental environment,comparing the effects of Hector and Gmapping,the latter algorithm was finally chosen as the mapping scheme of this thesis;Secondly,considering the slow path-finding speed of A* algorithm,the actual working environment of the robot,its own size and other attributes,the forced neighbor screening method in the jump point was improved.Based on the A* algorithm optimized by the jump point search,the J-A* global path planning algorithm was obtained,which improves the path-finding efficiently.Finally,the DWA local path planning algorithm was analyzed to ensure that the mobile platform can realize local obstacle avoidance on the global path.3.Motion planning of manipulator and an improved DG-RRT* algorithm of the path planning.Firstly,the polynomial interpolation problem of joint space trajectory planning and the interpolation problem of cartesian space trajectory planning were analyzed in detail.Secondly,in view of the slow convergence speed and redundant path of RRT* algorithm,an improved DG-RRT* algorithm was proposed,which makes use of the gravity of the target twice,thus ensuring the target orientation in the path-finding process,speeding up the convergence speed,pruning and smoothing the obtained path,and ensuring the stability of the manipulator operation.4.Experimental verification and simulation based on ROS system.Firstly,the software design of mobile manipulator in ROS system was completed.Secondly,experimental verification was carried out for the main research content of this thesis.The results show that the overall design scheme of the mobile manipulator is reasonable and feasible.