Hybrid Model Predictive Control with Physics-Informed Neural Network for Satellite Attitude Control
#satellite attitude control #model predictive control #physics-informed neural networks #system identification #computational efficiency #spacecraft dynamics
📌 Key Takeaways
- Hybrid MPC framework integrates physics‑informed neural networks for satellite attitude control.
- Accurate internal system models are critical for effective MPC, yet complex dynamics make traditional physics models hard to obtain.
- Learning‑based system identification offers a promising alternative to capture dynamics without exhaustive modeling.
- The approach targets reliability, predictive accuracy, and computational efficiency in spacecraft attitude control.
- Release date: February 2026 (arXiv:2602.15954v1).
📖 Full Retelling
Researchers report on a hybrid Model Predictive Control (MPC) framework that incorporates a Physics-Informed Neural Network to enhance satellite attitude control. The study, released on arXiv in February 2026, addresses the challenges faced by spacecraft with complex dynamics where accurate physics‑based models are difficult, time‑consuming, or computationally heavy to develop. By blending learning‑based system identification with physics constraints, the authors aim to improve the reliability and performance of MPC for satellite attitude regulation.
🏷️ Themes
Spacecraft dynamics, Model Predictive Control, Physics‐Informed Neural Networks, Learning‐Based System Identification, Attitude Control
Entity Intersection Graph
No entity connections available yet for this article.
Original Source
arXiv:2602.15954v1 Announce Type: cross
Abstract: Reliable spacecraft attitude control depends on accurate prediction of attitude dynamics, particularly when model-based strategies such as Model Predictive Control (MPC) are employed, where performance is limited by the quality of the internal system model. For spacecraft with complex dynamics, obtaining accurate physics-based models can be difficult, time-consuming, or computationally heavy. Learning-based system identification presents a compe
Read full article at source