Arne Rüting successfully completed his doctorate on the topic of "Realtime-capable model-predictive planning of time-optimized trajectories for kinematically redundant mechanism on industrial hardware".
Summary of the thesis:
This thesis deals with the model-predictive planning of time-optimized trajectories for kinematically redundant multi-axes mechanisms. First, basic terms are explained before the current state-of-science and technology with regard to path planning for kinematically redundant and non-redundant mechanisms as well as the execution of optimizations on industrial controllers is presented. Next, the model-predictive planning is explained. Hereby, the ambiguity when calculating the joint positions of a kinematically redundant mechanism is used to minimize the movement time between set points with regard to point-to-point movements.
Using a kinematically redundant hybrid mechanism, the approach is first validated with simulations. With regard to the increasing demands of more flexible manufacturing in context of batch size-1, real-time capable self-optimizing controls are necessary. Such controls adjust themselves to changing tasks during runtime and interact with other systems. Therefore, the focus is on the real-time execution of the approach on industrial controllers. This is a crucial point to achieve the required flexibility and industrial applicability, as well as to differ significantly from the state-of-the-art. For this purpose, the approach is implemented on the real system and the ability of a real-time execution with optimization times less than one millisecond is proven.
