L1 Adaptive Augmentation for Geometric Tracking Control of Quadrotors

MATLAB simulation by Khai Nguyen

The main program MainGeoL1_v1.m includes three main sessions: (1) setup, (2) main loop, and (3) visualization

Setup:
- This project is developed in MATLAB using object-oriented programming (OOP). This is the first time I try OOP in MATLAB and it works well.
- Classes:
  - Quadrotor: defines quadrotor parameters and dynamics
  - Geometry: defines methods associated to cordinates
  - Planner: defines simulation scenarios (initial states, desired trajectories and disturbances). Scenario details are provided below.
  - Controller: defines geometric controller (baseline)
  - Visualize: defines methods to draw figures and animation for visualization
  - L1AC: defines L1 Adaptive Control algorithms
Main loop:
- Step 1: Generate disturbances (time-continuous)
- Step 2: Sample Ts
  - Calculate current desired trajectories
  - Calculate baseline control signal (Geometric Control)
  - Calculate adaptive control signal (L1AC)
  - Calculate total control input
- Step 3: Apply control input to plant dynamics
- Step 4: Return to Step 1 until finishing
Visualization
- Draw figures on position tracking error, attitude tracking error, velocity tracking error, control input and disturbance estimate
- Make animation of 3D quadrotor motion
- Use MakeVideos.m to save gif

Planner(0, x) (x is a unused arbitrary value) presents a hovering quadrotor with one type of disturbance.
Planner(1, x) presents a quadrotor following a helix trajectory with disturbance dependent on x.
- x = 1: No disturbance
- x = 2, 3, 4: Different disturbances, see codes and the original paper for details
Planner(2, x) presents a quadrotor performing a flip with disturbance dependent on x.
- x = 1: No disturbance
- x = 2, 3, 4: Different disturbances, see codes and the original paper for details

Check out Simulation Results.docx

Please give me a ⭐ if you find my repo useful!