Learning Satellite Attitude Dynamics with Physics-Informed Normalising Flow

2025-08-11

AI Summary

To address the limitations of conventional model predictive control (MPC) in spacecraft attitude control—namely, its reliance on incomplete physical models—and the poor generalizability and stability of purely data-driven approaches, this paper proposes a physics-informed normalized flow learning framework. The method innovatively integrates physics-informed neural networks (PINNs) into a Real NVP architecture enhanced with self-attention mechanisms, trained on Basilisk-simulated attitude dynamics data and regularized by physics-based constraint losses. This enables high-fidelity, interpretable modeling of nonlinear attitude dynamics. Evaluated under unseen inputs, the framework reduces generalization error by 27.08%; when embedded within an MPC controller, it improves control accuracy by 42.86%. Moreover, it demonstrates superior noise robustness and closed-loop stability. Results validate the effectiveness and practicality of synergistically combining physical priors with invertible generative models for safety-critical aerospace control.

Technology Category

• Intelligent Robots: State Estimation
• Machine Learning: Calibration & Uncertainty Quantification
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Attitude control is a fundamental aspect of spacecraft operations. Model Predictive Control (MPC) has emerged as a powerful strategy for these tasks, relying on accurate models of the system dynamics to optimize control actions over a prediction horizon. In scenarios where physics models are incomplete, difficult to derive, or computationally expensive, machine learning offers a flexible alternative by learning the system behavior directly from data. However, purely data-driven models often struggle with generalization and stability, especially when applied to inputs outside their training domain. To address these limitations, we investigate the benefits of incorporating Physics-Informed Neural Networks (PINNs) into the learning of spacecraft attitude dynamics, comparing their performance with that of purely data-driven approaches. Using a Real-valued Non-Volume Preserving (Real NVP) neural network architecture with a self-attention mechanism, we trained several models on simulated data generated with the Basilisk simulator. Two training strategies were considered: a purely data-driven baseline and a physics-informed variant to improve robustness and stability. Our results demonstrate that the inclusion of physics-based information significantly enhances the performance in terms of the mean relative error of the best architectures found by 27.08%. These advantages are particularly evident when the learned models are integrated into an MPC framework, where PINN-based models consistently outperform their purely data-driven counterparts in terms of control accuracy and robustness, yielding improvements of up to 42.86% in performance stability error and increased robustness-to-noise.

Problem

• Improving spacecraft attitude control accuracy with physics-informed learning
• Addressing generalization issues in data-driven spacecraft dynamics models
• Enhancing MPC performance by integrating physics constraints into neural networks

Innovation

• Physics-Informed Neural Networks for spacecraft dynamics
• Real NVP with self-attention for attitude modeling
• Enhanced MPC performance with physics-based learning