🛰️ SDP-Based Satellite Attitude Control

Constrained spacecraft attitude control using Semidefinite Programming (SDP) with quaternion dynamics. The controller tracks a desired orientation while respecting geometric keep-in/keep-out cones and actuator limits. Simulations are in MATLAB/Simulink; optimization is solved via MOSEK through YALMIP or CVX.

🚀 Overview

• Method: SDP (MOSEK via YALMIP or CVX)
• States: Unit quaternions (q), body rates (\omega)
• Constraints: Keep-in/keep-out cones, bounds on (|\omega|) and torque (u), terminal state
• Env: MATLAB & Simulink

📂 Features

• Rigid-body (Euler) rotational dynamics + quaternion kinematics
• Exact cone constraints via SDP
• Plots for boresight on sphere, constraint margins, attitude/rate histories
• Baseline PD vs. SDP-guided tracking comparison

🧮 Dependencies

• MATLAB R2021b+ and Simulink
• Modeling: YALMIP or CVX
• Solver: MOSEK (installed, licensed, on MATLAB path)
  • Using CVX? Set cvx_solver mosek.

🛠️ How to run

1. Open the repo in MATLAB and set the current folder to the repo root.
2. Add paths: Home → Set Path → Add with Subfolders… → select repo root → Save.
3. Configure a scenario: edit Code/initialize_satellite_simulation.m (desired quaternion, torque/rate limits, cone angles).
4. Solve (SDP): run initialize_satellite_simulation (calls solveSDPControl.m). Results are written to Results/<run_name>.
5. Plot: run plot_SDPsim_results.m (or plot_SDP_sim_results.mlx) to save figures in Results/<run_name>.
6. (Optional) Simulink tracking: open CubeSat_SDP_PD_Attitude.slx, select SDP reference (or baseline PD), and Run.
   • What Simulink does: a quaternion-PD feedback regulator compares measured attitude to the SDP reference, calculates the error caused by the model uncertainities, and tries to minimise it, by appling control torque.

📜 License

MIT License

---

Made with ☕ and 🚀 by Shavy Kashyap