Sim-to-Real Transfer in Deep Reinforcement Learning for Bipedal Locomotion

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Bao, Lingfan, Peng, Tianhu, Zhou, Chengxu
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908639176949760
author Bao, Lingfan
Peng, Tianhu
Zhou, Chengxu
author_facet Bao, Lingfan
Peng, Tianhu
Zhou, Chengxu
contents This chapter addresses the critical challenge of simulation-to-reality (sim-to-real) transfer for deep reinforcement learning (DRL) in bipedal locomotion. After contextualizing the problem within various control architectures, we dissect the ``curse of simulation'' by analyzing the primary sources of sim-to-real gap: robot dynamics, contact modeling, state estimation, and numerical solvers. Building on this diagnosis, we structure the solutions around two complementary philosophies. The first is to shrink the gap through model-centric strategies that systematically improve the simulator's physical fidelity. The second is to harden the policy, a complementary approach that uses in-simulation robustness training and post-deployment adaptation to make the policy inherently resilient to model inaccuracies. The chapter concludes by synthesizing these philosophies into a strategic framework, providing a clear roadmap for developing and evaluating robust sim-to-real solutions.
format Preprint
id arxiv_https___arxiv_org_abs_2511_06465
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Sim-to-Real Transfer in Deep Reinforcement Learning for Bipedal Locomotion
Bao, Lingfan
Peng, Tianhu
Zhou, Chengxu
Robotics
This chapter addresses the critical challenge of simulation-to-reality (sim-to-real) transfer for deep reinforcement learning (DRL) in bipedal locomotion. After contextualizing the problem within various control architectures, we dissect the ``curse of simulation'' by analyzing the primary sources of sim-to-real gap: robot dynamics, contact modeling, state estimation, and numerical solvers. Building on this diagnosis, we structure the solutions around two complementary philosophies. The first is to shrink the gap through model-centric strategies that systematically improve the simulator's physical fidelity. The second is to harden the policy, a complementary approach that uses in-simulation robustness training and post-deployment adaptation to make the policy inherently resilient to model inaccuracies. The chapter concludes by synthesizing these philosophies into a strategic framework, providing a clear roadmap for developing and evaluating robust sim-to-real solutions.
title Sim-to-Real Transfer in Deep Reinforcement Learning for Bipedal Locomotion
topic Robotics
url https://arxiv.org/abs/2511.06465