Combining Simulations and Real-robot Experiments for Bayesian Optimization of Bipedal Gait Stabilization

TL;DR

This paper presents a Bayesian optimization approach that combines simulations and real-robot experiments to efficiently tune gait stabilization parameters for bipedal robots, reducing costs and hardware wear.

Contribution

It introduces a method that effectively integrates simulation data with real-world experiments for robot gait parameter optimization.

Findings

The approach reduces the number of real experiments needed.

It improves gait stabilization performance.

Demonstrated on the igus Humanoid Open Platform.

Abstract

Walking controllers often require parametrization which must be tuned according to some cost function. To estimate these parameters, simulations can be performed which are cheap but do not fully represent reality. Real-robot experiments, on the other hand, are more expensive and lead to hardware wear-off. In this paper, we propose an approach for combining simulations and real experiments to learn gait stabilization parameters. We use a Bayesian optimization method which selects the most informative points in parameter space to evaluate based on the entropy of the cost function to optimize. Experiments with the igus Humanoid Open Platform demonstrate the effectiveness of our approach.