WP1.2 Coordination Meeting June 1, 2020 - ibpsa/project1 GitHub Wiki
Meeting Report WP1.2 ‘Modelica library for MPC’
1. MEETING SUBJECT, DATE
Subject: WP1.2
Date: 01-06-2020
Location: Skype for Business
Minutes taken by: Lieve Helsen (KU Leuven)
2. PARTICIPANTS
3. AGENDA and REPORT
3.1. Approval/Modification Agenda
Approved
3.2. Status & further discussion where needed
a. Emulators:
4 models selected to start MPC testing
i. BESTEST Air (LBNL/KU Leuven):
merged to the master – READY! (Dave)
ii. BESTEST Hydronic (KU Leuven /LBNL):
ready for merging – ALMOST READY! (Filip)
iii. 8-Zone Residential Hydronic (Engie/SDU):
In pull request, update needed for zone identifiers (Javier).
iv. Multi-Zone Office Hybrid Simple (KU Leuven/Engie):
Managed to compile the model with JModelica, simulation with JModelica does not run (action Iago/Javier). Have a look at the solvers used.
However further progress has been made in other models too!
v. Single-Zone Commercial Building (SDU/LBNL):
Tao is processing the feedback from Dave.
vi. Multi-Zone Air-Based Prototype Complex (PNNL/KU Leuven):
Iago gave feedback, occupancy profile is missing. To be processed by Sen. No update.
vii. Flexible Research Platform (ORNL/SDU):
Yeonjin is working on the control part now 10 zones office with VAV system.
Emulators need to be put on BOP-TEST repository in the directory testcases.
Emulators overview:
b. KPIs - scenarios
Choice between TMYx (1946-2018) and TMY3 (1973-2005) weather files. We have a data reader for TMY3 in IBPSA. Both will be compared in the representative days/weeks selection (action PNNL). Sen No update.
c. BOPTEST:
i. Test case peer review document:
Update sent by Dave on May 20. No remarks.
ii. Upgrade to Python 3:
Dave is working on new version in Docker container that uses Python 3 and PyFMI (which can be installed easily), no JModelica-compiled image needed anymore; unit tests are going on now. We do not see the need to support Python 2 nor to provide backwards compatibility for it.
iii. Generation of new weather files based on historical data:
to be used for grey and black box model identification (training data that differ from test data). No further progress (Dave).
iv. Working group on weather prediction uncertainties (Aritz):
Autoregressive model based on measurements and Q-Q mapping have been tried. Behaviour of the standard deviation is being studied now to judge whether the method is appropriate, to be continued.
v. BS2021 Conference:
1. Contest? Target control or AI community rather than BS community (Jan will connect and make a first proposal for paper/presentation/workshop to make BOPTEST visible, https://www.climatechange.ai).
2. Workshop? Introducing BOPTEST in Summer 2021 (follow-up of paper at BS2019 in Rome), inspiration towards use cases such as: training building operators, assist teaching ... to be discussed in break-out session (all)
3. Paper? Rather go for a journal paper before BS2021.
d. MPC Library (IbpsaMpc):
i. Unit testing framework (Dave):
No update.
e. Dissemination:
Broaden the user space (e.g. to Machine Learning people). How to reach these people? Jan: make a first suggestion. Cfr OpenAIGym, to be combined with first proposal for paper/presentation/workshop (see 2.c.v.1).