This repository contains the code and resources for my university thesis, which focused on the simulation and implementation of an Acrobot system. The Acrobot is a two-link underactuated robotic arm, commonly used in control theory research. This project bridges theoretical simulations with practical hardware implementation, showcasing the swing-up and stabilization of the Acrobot using Linear-Quadratic Regulator (LQR) control.
Real_Acrobot_Simulation3.mp4
-
Acrobot Swing-Up and Balance Simulation:
- A detailed MATLAB/Simulink simulation of the Acrobot, featuring:
- The ability to modify system attributes, such as:
- Adding noise to the system to test robustness.
- Observing simulated data versus simulated encoder data and their differences.
- A DC motor model implemented in the simulation.
- Customizable starting positions for the Acrobot.
- The ability to modify system attributes, such as:
- Parameters for the simulation are derived from real-world measurements, but they can be adjusted as needed.
- A detailed MATLAB/Simulink simulation of the Acrobot, featuring:
-
Theoretical Double Pendulum Simulation:
- Simulates a theoretical double pendulum to explore its chaotic dynamics and provide insights into the physics and mathematics of such systems.
- Encoder Computing:
- Reads encoder values to measure the angles of the Acrobot's links in real-time.
- Servo Motor Control:
- Controls the torque applied to the Acrobot's actuator using a servo motor (replacing the DC motor used in the simulation).
- Combined Computing:
- Integrates encoder readings and motor control for a closed-loop system in hardware.
- Simulink Models: A modular and interactive simulation environment for testing and refining the Acrobot control system.
- MATLAB Scripts: Key functionalities include LQR design, energy calculations, and visualization.
- Hardware Control:
- Arduino code for real-time encoder data processing and servo motor actuation.
- Customization:
- Modify system parameters (e.g., masses, lengths, and moments of inertia) based on hardware measurements.
- Test different starting positions for the Acrobot.
To run the simulation, execute the main.m script in MATLAB. This will:
- Load the system parameters derived from real-world measurements.
- Linearize the system around the operating point.
- Perform LQR calculations.
- Simulate the Acrobot swing-up and balancing behavior in Simulink.
- Generate visualization plots for analysis.
- Observe both ideal simulated data and noisy encoder-like simulated data to analyze differences.
- Test and tune the system under various starting conditions and parameters.
- Parameters in the simulation are based on real-world measurements to ensure accuracy.
- While the simulation uses a DC motor model, the implementation uses a servo motor for torque control.
- The implementation closely follows the behavior predicted by the simulation.
For an in-depth explanation of the project, including theoretical background, implementation details, and results, refer to the thesis document: Allamvizsga_szakdolgozat.pdf.