In Intermediate Physics Laboratory (2025) we have a project to analyze a chaotic pendulum system. This will be done through tracking the angles of a physical set up, making a theoretical model, then comparing the results of the two.
Here is an image of the physical setup we are working with:
Figure 1. Physical Setup.
The MATLAB code given to us for tracking LEDs in a video was broken, old, and quite frankly bad, so we developed a new solution using openCV and other python libraries. See this README.
We also used Lagrangians to make a theoretical model, then we simulated the model. See this README.
If you are a future group doing this lab and hope to code something cool, consider forking this repository!
Here is a poster of our progress made for the first lab of IPL:
Figure 2. Project Poster!
The presentation in class went well! Here is the slides we used.
Our verification shows Arm 1 matches simulation well (±5°) while Arm 2 diverges significantly (±50°), suggesting our frictionless model needs refinement.
- Add viscous damping terms to the equations of motion
- Fit physical parameters (masses, lengths, damping coefficients) from video data using inverse optimization
- Implement a hybrid physics-ML model to capture residual dynamics (air resistance, joint compliance)
- Characterize chaos via Lyapunov exponents and Poincaré sections
Made by Adam Field and Christopher Pacheco.
Thank you Professor Noviello and TAs Drew and Holden for all the help.