Kinematic Analysis And Synthesis Of Kinematically Redundant Hybrid Parallel Robots PDF Download

This thesis investigates motion control methods and physical human robot interaction (pHRI) control strategies for two robotic systems, namely a kinematically redundant hybrid parallel robot (KRHPR) and a macro-mini system. The kinematic analysis, the dynamic modelling, as well as the control methods proposed in the thesis can be generalized for a class of robots with similar architecture. The thesis firstly introduces a novel kinematically redundant (6+3)-degree-of-freedom (DoF) spatial hybrid parallel robot with revolute actuators. The kinematic equations are developed and the singularities are examined. The translational and rotational workspace of the robot is then analysed. Also, a new mechanism is introduced to operate a gripper using the redundant DoFs. Thanks to the backdrivability of the robot, a controller - which can flexibly switch between two modes: position control and interaction control - is developed to demonstrate the potential use of this robot for physical interaction without using a force/torque sensor or joint torque sensors. Secondly, the motion control problem is investigated for a class of spatial kinematically redundant hybrid parallel robots. The kinematics are recalled and the dynamics are analysed. Based on this analysis, a proposed method referred to as hybrid control algorithm is proposed. It combines a simplified computed-torque controller, that operates in the joint space, with a Cartesian compensation, that operates in the task space of the robot. The stability of this approach is verified. Then, experiments are carried out on two example architectures. The results are examined and compared to those obtained with other methods to validate the effectiveness of the proposed approach. The motion control of a macro-mini system, which combines the hybrid parallel robot and a gantry system, is then investigated. The kinematics and the dynamics of the combined system are mainly analysed in the task space since it can be assumed that the position of the macro and the mini is stably determined by their own controllers. Motion control methods, namely mid-ranging control and Model Predictive Control, are generalized and adapted. Also, the combination of PI and the redundancy resolution is proposed. Each control method is implemented and used to perform the same trajectory. Afterwards, the control error is determined in order to compare the performance of the different methods. The physical human robot interaction is then studied for each of the robotic platforms mentioned above. On the KRHPR, a stiffness-damping control is specifically developed for pHRI applications. On the macro-mini system, the interaction method is also examined. The stability and the operational performance is analysed in detail. Experiments involving pHRI are then conducted and some demonstrations of potential applications are also presented. Finally, the conclusion summarizes the results obtained and discusses current limitations and potential future work.