This repository implements model predictive control using the PyPSA's power flow as the optimization backend in each iteration of the rolling horizon.
This is a work in progress and only contains essential features.
Control is based on the interplay of three main components: the pypsa.Network, containing the system configuration, multiple objects of the Prophet-type, responsible for returning time series to the pypsa components and the Controller, rolling snapshots forward in time while storing optimization outcomes.
The repository has minimal requirements and its intended to stay this way. To install the dependencies run
conda env create -f environment.yaml
Additionally, a linear solver is needed. For this, please refer to the PyPSA installation guide.
A minimal but illustrative example is presented in our readthedocs. Feel free to take a look!
I am always happy about feedback and for people to contribute, feel free to reach out at lukas.franken@ed.ac.uk