Paper published in a book (Scientific congresses, symposiums and conference proceedings)
Deep Reinforcement Learning based Continuous Control for Multicopter Systems
Manukyan, Anush; Olivares Mendez, Miguel Angel; Geist, Matthieu et al.
2019In International Conference on Control, Decision and Information CoDIT, Paris 23-26 April 2019
Peer reviewed
 

Files


Full Text
ContinousControl_AManukyan.pdf
Publisher postprint (309.54 kB)
Download

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Reinforcement Learning; UAV; TRPO
Abstract :
[en] In this paper we apply deep reinforcement learning techniques on a multicopter for learning a stable hovering task in a continuous action state environment. We present a framework based on OpenAI GYM, Gazebo and RotorS MAV simulator, utilized for successfully training different agents to perform various tasks. The deep reinforcement learning method used for the training is model-free, on-policy, actor-critic based algorithm called Trust Region Policy Optimization (TRPO). Two neural networks have been used as a nonlinear function approximators. Our experiments showed that such learning approach achieves successful results, and facilitates the process of controller design.
Research center :
Interdisciplinary Centre for Security, Reliability and Trust (SnT) > Automation & Robotics Research Group
Disciplines :
Computer science
Author, co-author :
Manukyan, Anush ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT)
Olivares Mendez, Miguel Angel ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT)
Geist, Matthieu
Voos, Holger  ;  University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > Engineering Research Unit
External co-authors :
yes
Language :
English
Title :
Deep Reinforcement Learning based Continuous Control for Multicopter Systems
Publication date :
26 April 2019
Event name :
International Conference on Control, Decision and Information CoDIT
Event place :
Paris, France
Event date :
from 23-04-2019 to 26-04-2019
Audience :
International
Main work title :
International Conference on Control, Decision and Information CoDIT, Paris 23-26 April 2019
Publisher :
IEEE
Peer reviewed :
Peer reviewed
Available on ORBilu :
since 24 January 2020

Statistics


Number of views
125 (11 by Unilu)
Number of downloads
474 (22 by Unilu)

Scopus citations®
 
2
Scopus citations®
without self-citations
1

Bibliography


Similar publications



Contact ORBilu