Acquisition of a biped walking pattern using a Poincare map

摘要

We propose a model-based reinforcement learning algorithm for biped walking in which the robot learns to appropriately place the swing leg. This decision is based on a learned model of the Poincare map of the periodic walking pattern. The model maps from a state at a single support phase and foot placement to a state at the next single support phase. We applied this approach to both a simulated robot model and an actual biped robot. We show that successful walking patterns are acquired.