-
Notifications
You must be signed in to change notification settings - Fork 0
2023 02 02 coordination
Date: February 2, 2023, 8:00 - 9:00 Pacific Times.
The purpose of this meeting is to coordinate the Modelica IBPSA Working Group.
Join from PC, Mac, Linux, iOS or Android: https://lbnl.zoom.us/my/mwetter
Or iPhone one-tap (US Toll): +14086380968,6614042296# or +16465588656,6614042296#
Or Telephone: Dial: +1 408 638 0968 (US Toll) or +1 646 558 8656 (US Toll) Meeting ID: 661 404 2296 International numbers available: https://lbnl.zoom.us/zoomconference?m=_h5BuJ686mPy3rWEyKf4NROxLaeOV5J8
Miscellaneous
PV collector, battery and electrical component models (Laura Maier)
Modular heat pump (Fabian Wüllhorst)
Fan efficiency based on Euler method (Hongxiang Fu)
Creation of reduced order building models (Alessandro Maccharini)
New state selection (Michael Wetter)
Equivalent borehole g functions (Massimo Cimmino)
Weekly schedule (Michael, Dre)
Progress on pull requests, https://github.com/ibpsa/modelica-ibpsa/pulls
Michael Wetter
Ettore Zanetti
Filip Jorissen
Javier Arroyo
Fabian Wullhorst
Christian Vering
H. Casper Fu
Jelger Jansen
Laura Maier
Saranya Anbarasu
Suggested new structure with 15 min presentation at beginning.
HP model presentation https://github.com/ibpsa/modelica-ibpsa/pull/1628 by Fabian.
Fabian presented HP model developed by Aachen (presentation available on repository). Few key points:
The HP model can use American or European data (Condeser outlet or inlet)
The model has been validated using data from Aachen
SDF library issue (commercial, Casella said that Modelica Standard library will come to make everything open source for multi dimensional tables)
Discussion:
Different name for the “BlacBoxStarter” it can use different types of models from black box (NN) to white physical model, to norm models. (No answers given as of today)
Michael suggested to check input 0-1 of the chiller models that’s just been updated on the master branch of the building library
Added preconfigured heat pumps for ease of use (can be used as a starting point to develop your own modular model)
General comments from Michael: example should not be too long to avoid running times that are too long because it will take a long time (t = 10^5) (in particular check with OPTIMICA or Open modelica, results diverge and becomes hard to crosscheck)
Q&A:
Q: Michael asked about state events generation:
A: Fabian said that he used the models in different scenarios and thresholds and it seems fine so far. Simulation time is also very reasonable (check publication on the model).
PV Laura close to pull request, she has some issues with cross checking the results (windows problem in getting reference results, in linux might work): https://github.com/ibpsa/modelica-ibpsa/issues/1575
https://github.com/ibpsa/modelica-ibpsa/pull/1628.
See presentation
Model was reviewed, Michael and Casper last check looks ok after Filip review. Waiting for test to pass and then we can move forward. https://github.com/ibpsa/modelica-ibpsa/pull/1646
Minor issue: Max function non differentiable. Solution:
Use smoothmax instead of max to make it differentiable
Michael still needs to make second review, close to be finished.
https://github.com/ibpsa/modelica-ibpsa/pull/1627
https://github.com/ibpsa/modelica-ibpsa/issues/1412
Michael work on model initialization? Difference of pressure instead of absolute pressure (in theory is better, in numerical simulation it seems inconclusive it depends on the models, scaling seems to work better) Comparison speed vs accuracy? (speed and if models actually run). Maybe use bounded attribute of real variables.
https://github.com/ibpsa/modelica-ibpsa/pull/1623
Just passed the test, it scales better with borefields with large numers Massimo implemented. So it will be merged.
Reviewers needed for Modelica 2023 https://2023.international.conference.modelica.org/. Joint paper from IBPSA Modelica working group?