| A mobile Manipulating Robot combining a mobile platform and robotic manipulator(s),is a new kind of robotic system with active manipulation ability.In spite of having complementary merits,this kind of configuration may pose many challenges to the motion planning and control algorithms.When an active manipulator is used in the system,the robot has redundant degrees of freedom and high spatial dimension of overall coordinated planning.When the robot is working dynamically,the offline planning is no longer applicable and online planning need be carried out,due to time-varying and uncertain information of environment.In addition,the movement of robots are not only constrained by its kinematics and dynamics,but also by non-linear constraints,such as environment(e.g.obstacle avoidance,movement restricted),which greatly increases the complexity of motion planning.However,the existing motion planning methods for mobile manipulating robots often treat the mobile platform and the manipulator(s)separately,so that the capability and efficiency of the robot cannot be fully exerted.In order to give full play to the advantages of the mobile manipulator(s)system and provide new ideas for the dynamic motion planning of the mobile manipulating robot,the motion planning method of the wheeled mobile manipulator(s)under the dynamic operation is studied,taking an Omni-directional Wheeled Mobile Manipulating Robot(WMMR)as the testbed in this thesis.Based on the composition of WMMR system,the kinematic model,mechanical model and the dynamic model of WMMR system have been established.The coupling effect between the components of the system is analyzed in detail.Then the model is simplified for controller design reasonably,which lays a foundation for dynamic motion planning.Also,the interaction characteristics between wheel-ground,vehicle-manipulator(s)and manipulator(s)-environment are analyzed to provide theoretical support for dynamic stability criteria of robots.In addition,minimum set parameter based optimal excitation trajectory is used to identify the dynamic parameters of WMMR system,and then the complete dynamic model is obtained,which provides a reference model for dynamic motion planning.Advantages and disadvantages of robot stability criteria in common use have been analyzed,and an improved tip-over moment stability criterion(ITOMSC)has been proposed to evaluate the dynamic stability of robots.Based on this criterion,the overturning laws of WMMR are analyzed,and an online dynamic tip-over avoidance method is proposed,which is verified by simulation and experiment.This algorithm is sensitive to centroid and also less computation,which provides stability criteria for WWMR dynamic motion planning.Based on this algorithm framework of rolling optimization,the specific implementation method of WMMR system with multiple physical constraints and high-dimensional dynamic motion planning is systematically studied.According to the dynamic operation characteristics of robots,the motion planning problem is divided into two parts including kinematics and dynamics.In addition,the high-dimensional dynamic motion planning at the kinematic level mainly solve the problem of discrete configuration task motion planning,which considers some robort's constraints(pose,velocity and acceleration of each joint of the robot),dynamic stability constraints and dynamic environment constraints.Design of this planner need four steps,including the Nonlinear Model Predictive Control(NMPC)model design,the cost function design,the weight matrix selection and the constraint addition.The solution of the planner is transformed into QP and NLP problems,which are solved by numerical optimization tools.The high-dimensional dynamic motion planning at the dynamic level mainly solve the problem of continuous trajectory constraint task motion planning,which considers a task that interacts with the environment.By using the purly rigid collision model,the planner is also designed through the above four steps,and then the motion planning of robot movement subject to trajectory and force constraint is carried out.The experimental platform of WMMR system is built.In addition,the proposed rolling optimization based motion planning algorithm framework is verified from the perspectives of kinematics and dynamics by the typical tasks,namely vehicle-manipulator(s)coordination under WMMR discrete configuration(dynamic transportation,dynamic grab/place)and vehicle-manipulator(s)coordination(door opening)under continuous trajectory and force constraint respectively.Some work and achievements in this thesis provide a new idea for solving the motion planning of redundant and ultra-redundant manipulating robots in multi-physical constraints,high-dimensional space and dynamic environment. |
PDF Full Text Request