batteryopt is a battery operation optimization tool developed by Jakub Szcześniak and implemented by Samuel Letellier-Duchesne. The objective is to minimize the annual electricity costs of a battery-integrated PV system using a Mixed-Integer Linear Program (MILP). The algorithm is implemented using the pyomo library opening up the model to a large array of solvers (e.g.: Gurobi, GLPK, etc.).
conda create --name batteryopt python=3.7 # tested with 3.7, 3.8 and 3.9
conda activate batteryoptgit clone https://github.com/MITSustainableDesignLab/batteryopt.git
cd batteryopt
python setup.py install
Type batteryopt --help to access the command line options
batteryopt outputs an Excel file with the model Variables for each time step of the year:
| t | tf | M | P_dmd | P_elec | P_pv | Buying | Charging | Discharging | E_s | P_charge | P_discharge | P_dmd_unmet | P_grid | P_pv_excess | P_pv_export | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 1 | nan | nan | 60536.5 | 0.0002624 | 0 | 1 | 0 | 0 | 20000 | -0 | -0 | 60536.5 | 60536.5 | 0 | 0 |
| 1 | 1 | 1 | nan | 60536.5 | 0.0002624 | 0 | 1 | 0 | 0 | 20000 | -0 | 0 | 60536.5 | 60536.5 | 0 | 0 |
| 2 | 1 | 1 | nan | 60536.5 | 0.0002624 | 0 | 1 | 0 | 0 | 20000 | -0 | 0 | 60536.5 | 60536.5 | 0 | 0 |
| 3 | 1 | 1 | nan | 60536.5 | 0.0002624 | 0 | 1 | 0 | 0 | 20000 | -0 | 0 | 60536.5 | 60536.5 | 0 | 0 |
| 4 | 1 | 1 | nan | 60536.5 | 0.0002624 | 0 | 1 | 0 | 0 | 20000 | -0 | 0 | 60536.5 | 60536.5 | 0 | 0 |
The column names are: