Spacecraft Dynamics – a GNC Digital Twin

A spacecraft Guidance, Navigation, and Control (GNC) simulation platform for studying orbital motion, attitude dynamics, sensing, estimation, actuation, feedback control, and autonomous maneuvering.

Project goals

1. Build a modular spacecraft digital twin from first principles.
2. Connect orbital and rigid-body dynamics to realistic sensors and actuators.
3. Implement state estimation and closed-loop attitude/orbit control.
4. Quantify controller and estimator performance under disturbances and uncertainty.
5. Produce reproducible experiments, visualizations, and a publication-quality technical contribution.

Core stack

Software: Python, NumPy, SciPy, pandas, Matplotlib, JupyterLab, pytest
Models: orbital mechanics, rigid-body rotational dynamics, sensor/actuator models, environmental disturbances
Methods: state-space modeling, numerical integration, PID/LQR control, Kalman filtering/EKF, Monte Carlo analysis

Repository map

Path	Purpose
notebooks/	Numbered investigations and demonstrations
src/spacecraft_dynamics/	Reusable simulation/GNC package
configs/	Spacecraft, orbit, sensor, actuator, and experiment parameters
tests/	Unit and regression tests
data/	Small generated/reference datasets only
figures/	Generated plots and diagrams
docs/	Theory notes and technical documentation
publication/	Essay/paper manuscript, figures, and reproducibility material
examples/	Small runnable end-to-end scenarios

Planned investigations

1. Two-body orbital propagation
2. Rigid-body attitude propagation
3. Reaction-wheel attitude control
4. Sensor simulation and measurement noise
5. Attitude estimation
6. Closed-loop pointing and stabilization
7. Environmental disturbances
8. Autonomous maneuver sequence
9. Monte Carlo robustness study
10. Publication experiment

Publication direction

The final contribution should make a narrow, defensible technical claim rather than simply describe the simulator. A strong direction is a reproducible comparison of attitude-estimation/control performance under realistic sensor noise, actuator limits, model uncertainty, and external disturbances.

See publication/OUTLINE.md.