Python-based closed-loop simulation of an Apollo-style lunar powered descent.
The goal was not to recreate Apollo perfectly, but to build a modular navigation–guidance–control architecture and explore how guidance laws, actuator limits, sensor imperfections, and mass depletion interact in a realistic descent problem.
- 2D lunar translational dynamics (polar coordinates)
- Continuous mass depletion from thrust
- LVLH frame navigation
- Polynomial boundary-condition guidance
- Throttle limits and gimbal rate limits
- Discrete GNC loop driving a continuous plant
- Altitude measurement with injected sensor noise
The code is intentionally separated:
sim/— truth dynamics (r, θ, m)gnc/navigation.py— LVLH state mapping with noisy altitude measurementgnc/guidance.py— acceleration command generationgnc/control.py— thrust + pitch allocation with limitstelemetry.py— logging and plotting
The structure mirrors real flight software partitioning rather than a single-script trajectory solver.
Outputs include:
- Altitude and vertical velocity
- Horizontal velocity
- Commanded vs actual thrust
- Commanded vs actual pitch
- Remaining propellant mass
- Radar (noisy) vs true altitude comparison
Termination reports:
- Time of flight
- Impact velocity
- Delta-V used
- Remaining propellant
- Frame discipline (operate GNC in LVLH, not polar)
- Solving cubic boundary-value guidance each cycle
- Handling actuator saturation and rate limits
- Mass depletion coupling into dynamics
- Separation between plant and control timing
- Effects of altitude sensor noise on guidance performance
- Telemetry-driven validation instead of eyeballing trajectories
- Translational dynamics only (no attitude dynamics)
- Single-axis sensor noise (altitude only)
- No full state estimator (no Kalman filter)
- No terrain model
- Simplified thrust vector geometry
-
NASA Manned Spacecraft Center, Apollo Lunar Descent and Ascent Trajectories
https://ntrs.nasa.gov/api/citations/19700024568/downloads/19700024568.pdf -
Mission Planning for Lunar Module Descent and Ascent
https://ntrs.nasa.gov/api/citations/19720018205/downloads/19720018205.pdf -
Apollo 11 Mission Report
https://sma.nasa.gov/SignificantIncidents/assets/a11_missionreport.pdf
Active development. Planned next steps:
- Add full state estimation (EKF)
- Monte Carlo dispersions with sensor noise
- Higher-order integration refinement
- Hardware-in-the-loop experimentation