Distributed Reinforcement Learning of a Six-Legged Robot to Walk

We present the principles of the reinforcement arning we use for the training of the walk of a hexapod robot. The originality of our approach lies in the fact that the training is distributed, each leg has to achieve its own goal. A gait appears as a emerging behavior and results from the ´self-coordination´ of the movements of the legs. We give the results of the simulation we have carried out and open prospects for our future work.