Prediction of the Spacecraft Position Relatively to the Focal Line of the Solar Gravitational Lens by Neural Network

In this paper, the problem of spacecraft navigation in the vicinity of the focal line of a solar gravitational lens is investigated. The proposed method uses a convolutional neural network for predicting spacecraft position by processing images of extended distant source. Firstly, the relations betw...

Full description

Saved in:
Bibliographic Details
Published in2023 10th International Conference on Recent Advances in Air and Space Technologies (RAST) pp. 1 - 4
Main Authors Korneev, Kirill, Perepukhov, Denis, Shirobokov, Maksim
Format Conference Proceeding
LanguageEnglish
Published IEEE 07.06.2023
Subjects
Convolutional neural networks
Estimation
Filtration
gravitational lensing
image processing
Kalman filters
Navigation
neural network
Neural networks
Space vehicles
spacecraft
Online AccessGet full text

Cover

More Information
Summary:In this paper, the problem of spacecraft navigation in the vicinity of the focal line of a solar gravitational lens is investigated. The proposed method uses a convolutional neural network for predicting spacecraft position by processing images of extended distant source. Firstly, the relations between the spacecraft's position and the source's position are given. Then, image modeling and preprocessing stages are described. With respect to specifics of the image, a convolutional neural network for direct position prediction is designed. Then, a Kalman filtration is applied for spacecraft's velocity estimation by processing neural networks' output. The numerical experiments show that the proposed navigation system can estimate position with a mean error of 31 thousand kilometers and velocity with a mean error of 1 meter per second.
DOI:10.1109/RAST57548.2023.10197903