LEARNING HOSE TRANSPORT CONTROL WITH Q-LEARNING

Non-rigid physical links introduce highly nonlinear dynamics in Multicomponent Robotic Systems (MCRS), which can hardly be solved analytically. In this paper, we propose the use of reinforcement learning methods to allow the agents learn by themselves how to deal with this kind of elements, as opposed to classical control schemes. In this paper we deal with the simplest case: only one hose segment and one robot at the tip of the hose. The task is to move the hose tip to an approximate position in the space. Learning is performed and tested using a hose-MCRS simulation environment developed by our group.