Piecewise Overactuation of Parallel Mechanisms Following Singular Trajectories: Modeling, Simulation and Control

Piecewise Overactuation of Parallel Mechanisms Following Singular Trajectories: Modeling, Simulation and Control
复制标题

DOI:
10.1007/s11044-004-2532-1
复制
发表时间:
2004-12
影响因子:
3.4
通讯作者:
L. Ganovski;P. Fisette;J. Samin
L. Ganovski;P. Fisette;J. Samin
中科院分区:
工程技术2区
文献类型:
--
作者:
L. Ganovski;P. Fisette;J. Samin

文献摘要

被引文献

相似文献

本方法旨在将冗余驱动原理用于并联机器人结构,以便以所谓的“机械优势”的方式跟踪复杂的轨迹,这意味着对于任何轨迹配置,执行器都可以承受重要的载荷(内部和/或外部)。简要回顾了用于建模的多体系统的相对坐标形式和求解闭环系统运动学约束的稳健技术。然后回顾了机器人可操作性的概念(以及它与速度和力椭球的关系)。作为所开发方法的第一步,提出了非过驱动系统执行器的合适位置,该准则基于上述理论概念。然后,根据可能的“客户”要求,在分段轨迹规划的基础上,对执行器的数量和位置进行了优化。静态和动态方面都会被处理,具体取决于应用程序。一旦确定了执行器的数目和位置,就提出了超定逆动力学的解决方案。最后,通过两个在合适的环境下建模、仿真和控制的多体并联结构实例验证了该方法的有效性。
The present approach aims at using the principle of redundant actuation for parallel robotic structures in order to follow complex trajectories in a so-called “mechanically advantageous” way, meaning that for any trajectory configuration the actuators can bear important loads (internal and/or external). Multibody system formalisms using relative coordinates and a robust technique to solve the closed-loop kinematic constraints, which are used for the purpose of modeling, are briefly reviewed. The notion of robot manipulability (and its relation to the velocity and force ellipsoids) is then recalled. As an initial step of the approach developed, suitable locations for the actuators of the non-overactuated system are proposed, the criteria being based on the theoretical concepts mentioned above. Then, according to some possible “customer” requirements, the number of actuators and their location are optimally specified on the basis of a piecewise trajectory planning. Both static and dynamic aspects are treated, depending on the application. Once the number and the locations of the actuators have been determined, a solution for the overdetermined inverse dynamics is proposed. Finally, the approach is validated via two examples of multibody parallel structures that are modeled, simulated and controlled in a suitable environment.