Cerebellar inspired sensorimotor learning and control for robotics (360G-Wellcome-105404_Z_14_Z)

The aim of my research is to test how feedforward and feedback mechanisms operate in synchrony in the cerebellum not only to aid motor learning, but also to be able to adapt movement continuously. My PhD will involve two interlinked lines of research: one experimental; the other computational. The experimental part will be based on electrophysiological recordings in the cerebellum of rats. The purpose of my electrophysiological investigations will be to look at differences in the frequencies recorded in different cerebellar modules. These will involve both single spikes and local field potential activity.Furthermore, to be able to relate such activation frequencies to feedback/feedforward models, the recordings will involve actual motor activation in the animals. Therefore, the recordings will be conducted during behavioural tasks involving goal-directed movements and will include perturbations generated either by peripheral limb stimulation or by magnetic force field. The computational part of my PhD will involve expanding the Cerebellar Adaptive Filter Model into one which incorporates prediction learning mechanisms. This new model will be implemented into a robotic arm, and the results of the robot experiments will be compared to those of the animal experiments.