I was asked to prepare an intro to MPC for the lab. So I did. It begins with some data exploration and then model development and fitting, and finally, we dive into controls. This is formatted to be an assignment. Running the Jupyter notebook won't produce most of the output unless you complete the statements. The purpose of this notebook is to introduce you to MPC and to develop some intuition for it.
To get started, install python 3 using anaconda 3 (download for windows). Install these which are likely not included by default:
- seaborn (plotting)
- bokeh (interactive plots, html output)
- statsmodels
- cvxpy (get version >1.0, optimization)
- scikit-learn
- scipy (should already be installed)
Have both attached files in the same directory. Fire up Jupyter notebook and open my notebook, grab a coffee and...
Good luck!
//vderm