Task and Domain Adaptive Reinforcement Learning for Robot Control

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liu, Yu Tang, Singh, Nilaksh, Ahmad, Aamir
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914952369930240
author Liu, Yu Tang
Singh, Nilaksh
Ahmad, Aamir
author_facet Liu, Yu Tang
Singh, Nilaksh
Ahmad, Aamir
contents Deep reinforcement learning (DRL) has shown remarkable success in simulation domains, yet its application in designing robot controllers remains limited, due to its single-task orientation and insufficient adaptability to environmental changes. To overcome these limitations, we present a novel adaptive agent that leverages transfer learning techniques to dynamically adapt policy in response to different tasks and environmental conditions. The approach is validated through the blimp control challenge, where multitasking capabilities and environmental adaptability are essential. The agent is trained using a custom, highly parallelized simulator built on IsaacGym. We perform zero-shot transfer to fly the blimp in the real world to solve various tasks. We share our code at https://github.com/robot-perception-group/adaptive_agent.
format Preprint
id arxiv_https___arxiv_org_abs_2404_18713
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Task and Domain Adaptive Reinforcement Learning for Robot Control
Liu, Yu Tang
Singh, Nilaksh
Ahmad, Aamir
Robotics
Artificial Intelligence
Systems and Control
Deep reinforcement learning (DRL) has shown remarkable success in simulation domains, yet its application in designing robot controllers remains limited, due to its single-task orientation and insufficient adaptability to environmental changes. To overcome these limitations, we present a novel adaptive agent that leverages transfer learning techniques to dynamically adapt policy in response to different tasks and environmental conditions. The approach is validated through the blimp control challenge, where multitasking capabilities and environmental adaptability are essential. The agent is trained using a custom, highly parallelized simulator built on IsaacGym. We perform zero-shot transfer to fly the blimp in the real world to solve various tasks. We share our code at https://github.com/robot-perception-group/adaptive_agent.
title Task and Domain Adaptive Reinforcement Learning for Robot Control
topic Robotics
Artificial Intelligence
Systems and Control
url https://arxiv.org/abs/2404.18713