-
Notifications
You must be signed in to change notification settings - Fork 84
Functional Requirements
Physics to be modeled include:
- Pressure driven flow, using a quadratic relation with linearization near zero.
- Air damper models and valve models with leakage flow.
- Dynamics of sensors, approximated by a first order element, and of actuators.
- Pumps and fans with the pressure-flow relation based on parameterizable performance curves.
- Air with CO2 and humidity concentration, and water, based on the Modelica.Media library.
- Transport delay in fluid flow networks. In a first version, this will be approximated by a first order delay; with a plug flow model being added later as needed.
- Adiabatic cooling using a cooling tower or an air humidifier.
- Cooling coil with water vapor condensation and heating coil.
- Chiller with variable frequency drive.
The models are based on the Modelica.Fluid library.
Currently, out of scope are modeling of:
- Refrigerants, including two-phase flow.
- Package transmission in communication networks. See also the Modelica library presented in Simulation of Distributed Automation Systems in Modelica by Wagner, Liu and Frey.
Different model idealizations will be used: Some models will resolve dynamics with characteristic time constants in the order of seconds, whereas other models will only take slower dynamics into account, or in the limiting case, be steady-state models. What model will be developed depends on the use case and the requirements needed for the particular use case (such as for thermal energy management or design of local feedback control).
Where applicable, a model user should be able to use a dynamic or a steady-state model.
To be able to model a building's thermal response in EnergyPlus and to model supervisory controls in MATLAB/Simulink or in Ptolemy II, interfaces need to allow linking Modelica models to other simulators using
- Functional Mockup Units,
- the Building Controls Virtual Test Bed, and
- direct coupling through Python 2.7 embedded in Modelica.
To automate post processing, we will provide Python scripts that can extract time series from the Dymola generated output file and from EnergyPlus generated CSV files.
To test component models and to provide to other users examples for possible model use, for each component model there will be a test model that illustrates the component use and that can be automatically run from a Dymola script for unit testing.
To use TMY3 weather data, a script will be developed that converts TMY3 weather data, obtainable from the EnergyPlus weather data web site, to a format that can be efficiently processed by Modelica.
For computational efficiency, model equations shall were possible be differentiable and have a continuous first derivative.
Whenever possible, a Modelica tool should not have to do numerical differentiation.
All models will exclusively use SI units.