Asteroid gravitational field calculation via GeodesyNets with quadratic layers

Asteroid gravity inversion is a method used to infer the internal density distribution of small celestial bodies through external gravitational measurements. In recent years, gravity inversion techniques based on neural networks have seen rapid development, such as GeodesyNet, which utilizes a fully...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhitao Fu, Weitong Li, Shanhong Liu
Format: Article
Language:English
Published: Taylor & Francis Group 2025-06-01
Series:Geo-spatial Information Science
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/10095020.2025.2506762
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849471031955161088
author Zhitao Fu
Weitong Li
Shanhong Liu
author_facet Zhitao Fu
Weitong Li
Shanhong Liu
author_sort Zhitao Fu
collection DOAJ
description Asteroid gravity inversion is a method used to infer the internal density distribution of small celestial bodies through external gravitational measurements. In recent years, gravity inversion techniques based on neural networks have seen rapid development, such as GeodesyNet, which utilizes a fully connected SIREN neural network. The traditional SIREN neural network performs well in capturing low-frequency features, such as regular shapes or large-scale density distributions. However, it has limitations in learning high frequency features when processing irregular shapes or abrupt changes in the inner density field. This study introduces an optimized architecture that synergistically integrates SIREN networks with quadratic neurons to augment the SIREN network’s capability in capturing high-frequency gravitational features of asteroids, which seeks to improve the accuracy and stability of inversion processes within the GeodesyNet framework. Eros, Itokawa, 67P/Churyumov-Gerasimenko, and Bennu are selected for testing and evaluation. The results indicate that the density distributions computed by this combined network are more in line with the actual conditions. The maximum relative root mean square error (relRMSE) for homogeneous test models decreased by 6.1% for 67P, and through differential training, the maximum relRMSE for non-homogeneous asteroid surfaces was reduced by 3.0% for Bennu. Moreover, the improvement from GeodesyNetsQ is larger when fewer iterations, and thus fewer data points are used, which will be more helpful to be applied in practical asteroid missions with limited measurements.
format Article
id doaj-art-2a68529794a04b3797a55f5c8acf4cd2
institution Kabale University
issn 1009-5020
1993-5153
language English
publishDate 2025-06-01
publisher Taylor & Francis Group
record_format Article
series Geo-spatial Information Science
spelling doaj-art-2a68529794a04b3797a55f5c8acf4cd22025-08-20T03:24:57ZengTaylor & Francis GroupGeo-spatial Information Science1009-50201993-51532025-06-0111310.1080/10095020.2025.2506762Asteroid gravitational field calculation via GeodesyNets with quadratic layersZhitao Fu0Weitong Li1Shanhong Liu2Faculty of Land Resources Engineering, Kunming University of Science and Technology, Kunming, ChinaFaculty of Land Resources Engineering, Kunming University of Science and Technology, Kunming, ChinaBeijing Aerospace Control Center, Beijing, ChinaAsteroid gravity inversion is a method used to infer the internal density distribution of small celestial bodies through external gravitational measurements. In recent years, gravity inversion techniques based on neural networks have seen rapid development, such as GeodesyNet, which utilizes a fully connected SIREN neural network. The traditional SIREN neural network performs well in capturing low-frequency features, such as regular shapes or large-scale density distributions. However, it has limitations in learning high frequency features when processing irregular shapes or abrupt changes in the inner density field. This study introduces an optimized architecture that synergistically integrates SIREN networks with quadratic neurons to augment the SIREN network’s capability in capturing high-frequency gravitational features of asteroids, which seeks to improve the accuracy and stability of inversion processes within the GeodesyNet framework. Eros, Itokawa, 67P/Churyumov-Gerasimenko, and Bennu are selected for testing and evaluation. The results indicate that the density distributions computed by this combined network are more in line with the actual conditions. The maximum relative root mean square error (relRMSE) for homogeneous test models decreased by 6.1% for 67P, and through differential training, the maximum relRMSE for non-homogeneous asteroid surfaces was reduced by 3.0% for Bennu. Moreover, the improvement from GeodesyNetsQ is larger when fewer iterations, and thus fewer data points are used, which will be more helpful to be applied in practical asteroid missions with limited measurements.https://www.tandfonline.com/doi/10.1080/10095020.2025.2506762Asteroidgravity inversionneural network
spellingShingle Zhitao Fu
Weitong Li
Shanhong Liu
Asteroid gravitational field calculation via GeodesyNets with quadratic layers
Geo-spatial Information Science
Asteroid
gravity inversion
neural network
title Asteroid gravitational field calculation via GeodesyNets with quadratic layers
title_full Asteroid gravitational field calculation via GeodesyNets with quadratic layers
title_fullStr Asteroid gravitational field calculation via GeodesyNets with quadratic layers
title_full_unstemmed Asteroid gravitational field calculation via GeodesyNets with quadratic layers
title_short Asteroid gravitational field calculation via GeodesyNets with quadratic layers
title_sort asteroid gravitational field calculation via geodesynets with quadratic layers
topic Asteroid
gravity inversion
neural network
url https://www.tandfonline.com/doi/10.1080/10095020.2025.2506762
work_keys_str_mv AT zhitaofu asteroidgravitationalfieldcalculationviageodesynetswithquadraticlayers
AT weitongli asteroidgravitationalfieldcalculationviageodesynetswithquadraticlayers
AT shanhongliu asteroidgravitationalfieldcalculationviageodesynetswithquadraticlayers