Model predictive control algorithms applied to various dynamic systems, developed in Python. Using the do-mpc and casadi package in Python. Systems implemented:
- Spaceship
- Kinematic bicycle model
models: Contains subfolders for each model. Model directory will have a description of the system, animated results and plots, and details for how to modify trajectories and tune parameters.template: Contains template code to implement MPC on new systems withdo-mpcpackage.visualizer [WIP]: Folder containing Unity assets, resources, and scripts to visualize results from MPC control.
A conda environment running Python 3.x is recommended with the following packages
- numpy
- CasADi
- matplotlib
The details for installing do-mpc can be found here: https://www.do-mpc.com/en/latest/installation.html
It is recommended to install the HSL MA27 solver for faster computation, which can be found here: http://www.hsl.rl.ac.uk/ipopt/