To create a representative heat pump model, performance data
from multiple commercial air-to-water heat pumps were first
digitized. These datasets included information such as coefficient
of performance (COP) and heating power across a range
of outdoor air temperatures and fixed supply temperatures. For each
individual heat pump, regression models were fitted using
polynomial functions. These regressions captured the variation of
COP and heating power with outdoor temperature for
each fixed supply temperature level. In a subsequent step, these
individual regressions were used to derive a generalized model. For
each selected supply temperature (e.g., 35 °C, 45°C, 55°C, and
60°C), a second polynomial regression was performed across all
previously generated curves.
The resulting regression coefficients were then used to define a
Modelica record representing an average heat pump. This record
provides COP and heating power as continuous functions
of the outdoor temperature and is intended for use in system-level
simulations where a generic heat pump behavior is required.