Overview of Meta-Reinforcement Learning Methods for Autonomous Landing Guidance

Andrea Scorsoglio, Luca Ghilardi, Roberto Furfaro

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents a vision-based method for autonomous planetary powered descent and landing using meta-reinforcement learning. The goal is map observations to thrust commands directly using a deep neural network. Two iterations of the method are presented. First the 3-degrees-of-freedom powered descent pinpoint landing is solved using images, altitude and vertical rate as inputs. Then the problem of landing site selection, solved using a secondary neural network that performs semantic segmentation, is introduced. The final model is capable of autonomously select a landing area and guide the spacecraft to the designated landing site using only images as input.

Original languageEnglish (US)
Title of host publicationThe Use of Artificial Intelligence for Space Applications - Workshop at the 2022 International Conference on Applied Intelligence and Informatics
EditorsCosimo Ieracitano, Nadia Mammone, Marco Di Clemente, Mufti Mahmud, Roberto Furfaro, Francesco Carlo Morabito
PublisherSpringer Science and Business Media Deutschland GmbH
Pages49-63
Number of pages15
ISBN (Print)9783031257544
DOIs
StatePublished - 2023
Event2nd International Conference on Applied Intelligence and Informatics , AII 2022 - Reggio Calabria, Italy
Duration: Sep 1 2022Sep 3 2022

Publication series

NameStudies in Computational Intelligence
Volume1088

Conference

Conference2nd International Conference on Applied Intelligence and Informatics , AII 2022
Country/TerritoryItaly
CityReggio Calabria
Period9/1/229/3/22

Keywords

  • Meta-reinforcement learning
  • Spacecraft guidance

ASJC Scopus subject areas

  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'Overview of Meta-Reinforcement Learning Methods for Autonomous Landing Guidance'. Together they form a unique fingerprint.

Cite this