Download the package of jEPlus+EA including the calibration project (updated on 25/11/2015).
Monthly energy and demand data tables from SQLite output.
Use Python to calculated the Normalized Mean Bias Error (NMBE) and the Max relative error (MAXE)
MAXE values for energy and demand are used as constraints.
NMBE values for energy and demand are used as objectives.
MAXE values within 20% are deemed as success, shown as the highlighted region of chart below:
NMBE values should be as close to 0 as possible.