A cognitive robotics approach to the human reaching process: simulating goal-directed reaching and choice-reaching tasks with a closed-loop model

摘要

In this thesis a robotics-based model for simulating reaching experiments is presented. First, I will focus on simulating visual attention in choice-reaching tasks where participants are asked to make rapid reach movements towards a target that is presented with distractors. These studies found that distracters can alter the movement trajectory substantially (e.g. Song & Nakayama, 2008). In order to understand the complex behaviour in such tasks I will follow a robotics based approach and simulate numerous choice-reaching tasks.ududThe second part of this thesis deals with phenomena of goal-directed reaching such as the speed-accuracy trade-off of Fitts' law (Fitts, 1954). In the simulations I will demonstrate the human-like goal-directed reaching behaviour of my model. Moreover, due to its feedback-based architecture my model offers an alternative explanation of the goal-directed reaching process.ududMy model utilises a real-world LEGO robot and required the implementation of multiple underlying processes and the handling of noise. To link the model with the control of the robot I applied the neurological plausible framework of the dynamic field theory of Erlhagen and Schoener (2002). The successful simulations demonstrated that my model can give further insight in the nature of goal-directed reaching and the choice-reaching tasks.ud