Logo image
Back
Dynamic Stabilization of Unstable Limit Cycles in Passive Walking via Deep Reinforcement Learning
Conference proceeding

Dynamic Stabilization of Unstable Limit Cycles in Passive Walking via Deep Reinforcement Learning

Fateme Ranjbaran, Mahyar Naraghi, Vahid Ehteshami Bojnourdi and Ali Tehrani Safa
11th RSI International Conference on Robotics and Mechatronics, ICRoM 2023, pp 586-591
2023
DOI: https://doi.org/10.1109/ICRoM60803.2023.10412468

Abstract

Deep Reinforcement Learning Passive dynamic walking Reversible Switching Surface Soft Actor-Critic
In this research, Deep Reinforcement Learning (DRL) is employed to develop an ankle controller for a passive biped walker with round feet. The selected DRL technique is Soft Actor-Critic. A reward function is formulated and tailored to the specific requirements of the RL framework. Notably, this controller leverages passive dynamic Basin of Attraction data as initial conditions for the training episodic resets, which is an inherent characteristic of the model. A custom Gym environment is constructed to facilitate the testing of reinforcement learning algorithms. The findings are validated against a controller based on the Reversible Switching Surface (RSS) technique. The investigation demonstrates that the trained controller exhibits stable and robust gaits, achieving results appropriate to the optimal RSS controller gains without the need for additional sensitivity analysis.

Metrics

1 Record Views

Details

Logo image