Abstract

We have been developing a new reinforcement learning called BRL, which is proved to be especially effective to multi-robot systems(MRS). In this paper, we apply extended BRL to control three robots with two omni wheels. In this system, robots need to deal with uncertain state transition derived from passive movement by omni wheels. This uncertainty means robots have to segment their state and action spaces carefully. As a result of the experiment, robots acquired a necessary action for the task achievement.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call