Dynamic analysis and distributed control of the Tetrobot modular reconfigurable parallel robotic system.

摘要

Reconfigurable robotic systems can be adapted to different tasks or environments by reorganizing their mechanical configurations. Such systems have many redundant degrees of freedom in order to meet the combined demands of strength, rigidity, workspace kinematics, reconfigurability, and fault tolerance. In order to implement these new generations of robotic system, new approaches must be considered for design, analysis, and control.; This thesis presents an efficient distributed computational scheme which computes the kinematics, dynamics, redundancy resolution, and control inputs for real-time application to the control of the Tetrobot modular reconfigurable robots. The entire system is decomposed into subsystems based on a modular approach and Newton's equations of motion are derived and implemented using a recursive propagation algorithm. Two different dynamic resolution of redundancy schemes, the centralized Jacobian method and the distributed virtual force method, are proposed to optimize the actuating forces. Distributed dynamic control algorithms provide an efficient modular implementation of the control architecture for a large family of configurations. Simulation results are provided to demonstrate the feasibility of the proposed distributed scheme and compared to a centralized scheme for different configurations of the Tetrobots such as a serial chain manipulator, a six-legged walker, and an icosahedral rolling Tetrobot.; Complexity and stability are also important issues in dealing with large scale modular robotic systems because the computational burden and the error caused by time delay grow significantly as the number of modules increases. The computational complexity of the Tetrobot is presented and compares the computation time between the centralized and distributed algorithms. The stability analysis and time delay dependent criteria of time-delay systems are presented for a linear system with N degrees of freedom and a simulation study is performed to investigate the computation errors and stability of the distributed control algorithm.; Finally, the active rolling locomotion which describes the dynamic modeling, locomotion planning, and control is studied to provide a means of mobility to the Tetrobot modular robot. The motion is described by the path profiles of controlled nodes, the tipping criteria and dynamic tipping motion, and an impact-reaction model of contact with the ground. These phases of motion are described using Newton-Euler dynamic equations and the principle of conservation of angular momentum. A two-phase planning and switching control sequence is introduced to achieve stable and reliable motion of a Tetrobot modular robot. The resulting models are useful to analyze and control both intentional rolling as a new mode of mobility as well as the avoidance of unintentional tipping and rolling during task execution.