This thesis presents novel insights and solutions for the planning and control of robotic manipulation tasks with non-negligible dynamics. The central goal is to equip robots with advanced sensory and motor skills. Hereby, it represents an important contribution to the development of flexible and versatile robotic systems.
The thesis follows a model-based approach and discusses state-of-the-art models of the system and its environment. Methods for optimal motion planning and for hybrid control of dynamic manipulation tasks are presented and applied to a number of case studies. Specific challenges with respect to environment perception are addressed that are characteristic for dynamic object manipulation. In order to improve the system performance, a compliant end effector design is proposed and evaluated.
The effectiveness of the employed methods and the overall framework is demonstrated in a number of simulations and experiments.
«
This thesis presents novel insights and solutions for the planning and control of robotic manipulation tasks with non-negligible dynamics. The central goal is to equip robots with advanced sensory and motor skills. Hereby, it represents an important contribution to the development of flexible and versatile robotic systems.
The thesis follows a model-based approach and discusses state-of-the-art models of the system and its environment. Methods for optimal motion planning and for hybrid control...
»