2019 12 04_meeting - lbl-srg/soep GitHub Wiki

SOEP Coordination meeting

Date: December 4, 2019, 8:15am-9am Pacific Times

The purpose of this meeting is to coordinate the Modelica integration for SOEP.

Meeting information

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

Agenda

  1. Previous action items.
  2. Michael: SimBuild paper.
  3. Kyle: NREL/Modelon subcontract update.
  4. Kyle, Michael: Refactoring room air heat and mass balance.
  5. Kyle, Michael: EnergyPlus I/O variables.
  6. Stuart: FMI-QSS.
  7. Kyle, Michael: Optimica license protection.
  8. Jianjun: JSON representation for OS front-end to Modelica
  9. Other items

Participants

Yanfei Li
Jianjun Hu
Antoine Gautier
Stuart Mentzer
Amir Roth
Bo Shen
Hubertus Tummescheit
Michael Wetter

Minutes

SimBuild paper

Paper is developed at https://bitbucket.org/berkeleylab/srg-soep-papers/src/master/, see https://bitbucket.org/berkeleylab/srg-soep-papers/src/master/2020-simbuild/1_manuscript/

Review of current version.

Contributions due by the end of this week, the earlier the better. Final submission to conference due Fr Dec 13.

Stuart to try to figure out the event dependencies and simplifications for SimpleHouse. See also Optimica guides:

So, one flag for you is generate_mof_files, which will give you a flat Modelica model for finding all event-generating expressions There is a lot of info in Chapter 15 of the Optimica users guide about smoothness. There are various options on how to find everhting non-smooth, i.e. that generates switches disable_smooth_events set to true, then smooth will not generate events event_indicator_scaling

If it won't be practical, resort to example from BS'15 paper for which Michael can do a Modelica implementation.

Numerical investigations

No discussion.

No work conducted.

Refactoring room air balance

Branch issue1129_energyPlus_zone works for multiple buildings for JModelica, Dymola and OpenModelica. This has the FMI interface now.

Issues are at https://github.com/lbl-srg/modelica-buildings/issues?q=is%3Aissue+is%3Aopen+label%3Aspawn

Next, work on

  • Kyle to implement EnergyPlus output variables (segfault).
  • Michael to implement schedules and EMS actuators (need to have outputs working first).
  • Kyle to implement schedules and EMS actuators.
  • Kyle to make an FMU that does not compress the files.
  • Kyle to look into generating all outputs in the folder tmp-simulation-TwoIdenticalZones
  • Kyle to look into numerical instability (walls may be initialized to wrong temperature). See for example Experimental/EnergyPlus/Examples/VAVReheatRefBldgSmallOffice/FloorOpenLoop or Experimental/EnergyPlus/Validation/OneZone and set the initial room temperature to 10 degC.
  • Michael and Kyle to work out how to filter level of info/error/fatal errors. See code near

const size_t nCat = fmi2_import_get_log_categories_num(bui->fmu); * Installer package for all three platforms.

Contract:

  1. All zones in the idf file must have a zone model in Modelica. Otherwise the simulation stops with an error.
  2. If there is an HVAC system in the idf file, then EnergyPlus issues a warning, the EnergyPlus HVAC system is not simulated, but the coupled EnergyPlus/Modelica simulation proceeds.
  3. For the EnergyPlus envelope, either the CTF transfer function or the finite difference method can be used.

FMI-QSS

No discussion.

Stuart to work on updating 3rd order QSS with numerical differentiation.

Next will be work on performance estimation assuming we would have automatic differentiation.

Michael to set up simple control example with time sampling, PI and hysteresis. Put example to https://github.com/NREL/SOEP-QSS-Test

CI testing (funnel)

No discussions.

Next:

  • Work on experiment specification.

Optimica license protection

No discussion.

Subcontract between Modelon and NREL still not in place. Hubertus will have a call with Luigi.

Potential paths: Check diff to frozen library. Allow n models to be changed. Upgrade path to full OCT version (with access to cloud version in an integrated way).

Need to define the different use cases and work flows.

Use of python API makes protection much easier, is sufficient unless interactive mode is needed. Java API would be needed for interactive mode.

JSON representation for OS front-end to Modelica

No discussion.

Jianjun to work on modelica-json parsing in Modelica rather than CDL mode.

Ongoing is dealing with replaceable models.

Other items

None.

New action items

Michael to set up simple control example for QSS with time sampling, PI and hysteresis.

See next steps under "Refactoring room air balance".

Previous action items

Michael will look into making dp of pressure drop elements a state rather than p using stateSelect.

Agnes to get back regarding debug information for user for nonlinear solver.

⚠️ **GitHub.com Fallback** ⚠️