Learning Linear Policies for Robust Bipedal Locomotion on Terrains with Varying Slopes
This provides a lightweight control framework for bipedal robots, enabling robust locomotion on varying slopes, though it is incremental as it builds on existing methods like ARS.
The paper tackles the problem of controlling bipedal walking robots on sloped terrains by learning a single linear feedback policy using Augmented Random Search, achieving robust walking on slopes up to 20 degrees and recovery from pushes up to 120 N in simulation.
In this paper, with a view toward deployment of light-weight control frameworks for bipedal walking robots, we realize end-foot trajectories that are shaped by a single linear feedback policy. We learn this policy via a model-free and a gradient-free learning algorithm, Augmented Random Search (ARS), in the two robot platforms Rabbit and Digit. Our contributions are two-fold: a) By using torso and support plane orientation as inputs, we achieve robust walking on slopes of up to 20 degrees in simulation. b) We demonstrate additional behaviors like walking backwards, stepping-in-place, and recovery from external pushes of up to 120 N. The end result is a robust and a fast feedback control law for bipedal walking on terrains with varying slopes. Towards the end, we also provide preliminary results of hardware transfer to Digit.