Loading...

Satellite pose estimation using Earth radiation modeled by artificial neural networks

Nasihati Gourabi, F ; Sharif University of Technology | 2022

63 Viewed
  1. Type of Document: Article
  2. DOI: 10.1016/j.asr.2022.07.009
  3. Publisher: Elsevier Ltd , 2022
  4. Abstract:
  5. The thermal energy received by each surface of an Earth-orbiting satellite strongly depends on its position and orientation. In this sense, simultaneous orbit and attitude estimation (SOAE) using the surface temperature data has been focused in the present study. The Earth infrared (IR) radiation and the Earth's top-of-atmosphere (TOA) albedo are two key sources of radiation affecting the satellite surface temperature rate. The Earth's radiation information has been monitored for the past two decades by the Clouds and the Earth's Radiant Energy System (CERES) project, producing a comprehensive set of Earth radiation budget (ERB) data for climate, weather and applied science research. The current study utilizes the ERB data to forecast the IR radiation flux (IRF) and the TOA albedo factor (AF) via an artificial neural network (ANN). In this respect, the satellite longitude, latitude and time constitute the input vector of the learning set, while the AF and IRF are considered as the output. The ANN is then employed to model the satellite surface temperature rate in the radiation-based SOAE process. The feasibility of the proposed pose estimation approach has been addressed and verified via Monte Carlo simulation. Obtained results demonstrate suitability of the algorithm in sunlight intervals with good accuracy. © 2022 COSPAR
  6. Keywords:
  7. Artificial neural network ; Attitude estimation ; Atmospheric temperature ; Budget control ; Infrared radiation ; Intelligent systems ; Monte Carlo methods ; Orbits ; Satellites ; Solar radiation ; Surface properties ; Albedo radiation ; Budget data ; Earth radiation ; Earth radiation Budget ; Orbit estimation ; Pose-estimation ; Radiation fluxes ; Surface temperatures ; Top of atmospheres ; Neural networks
  8. Source: Advances in Space Research ; Volume 70, Issue 8 , 2022 , Pages 2195-2207 ; 02731177 (ISSN)
  9. URL: https://www.sciencedirect.com/science/article/abs/pii/S027311772200583X