Robust Attitude Estimation for Nanosatellites with Deployable Panels Shadowing the Sun Sensors


Esit M., Söken H. E., Cordova R., Hiraki K.

2025 IAF Astrodynamics Symposium at the 76th International Astronautical Congress, IAC 2025, Sydney, Avustralya, 29 Eylül - 03 Ekim 2025, cilt.1-F219391, ss.39-46, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Cilt numarası: 1-F219391
  • Doi Numarası: 10.52202/083087-0004
  • Basıldığı Şehir: Sydney
  • Basıldığı Ülke: Avustralya
  • Sayfa Sayıları: ss.39-46
  • Anahtar Kelimeler: adaptive covariance, attitude estimation, coarse sun sensor, robust filtering, self-shadowing, sensor malfunction
  • Orta Doğu Teknik Üniversitesi Adresli: Evet

Özet

Attitude estimation plays an essential role in nanosatellites, whether for pointing accuracy or mission analysis where attitude information is required. Typically, low-cost attitude determination systems rely on gyroscopes, magnetometers, and coarse sun sensors to estimate attitude and angular rates. For nanosatellites equipped with deployable panels, the coarse sun sensors, which are not optimally located on the spacecraft, may experience self-shadowing effects, leading to large attitude estimation errors. This study was motivated by the Kyutech 3U Leopard CubeSat project, which features deployable panels and coarse sun sensors mounted on the body panels. Simulations incorporating a shadow model revealed significant degradation in attitude estimation during self-shadowing events at specific mission phases. To address this issue, this study investigates various enhancement techniques, including adaptive covariance matrix adjustment and a second-stage estimator, to improve robustness and maintain estimation accuracy despite sensor degradation. The proposed attitude filtering strategies were evaluated under different scenarios, including tumbling motions with varying angular rates. The results demonstrate that some of the proposed methods effectively maintain attitude estimation performance during self-shadowing events while achieving accuracy comparable to cases without shadowing, ensuring reliable nanosatellite attitude determination in real-world operations.