-
Notifications
You must be signed in to change notification settings - Fork 39
WP1.2 Coordination Meeting March 21, 2019
Meeting Report WP1.2 ‘Modelica library for MPC’
Subject: WP1.2
Date: 21-03-2019
Location: Skype for Business
Minutes taken by: Lieve Helsen (KU Leuven)
Company/Organisation Participant
KU Leuven Lieve Helsen (WP Leader)
KU Leuven Filip Jorissen
KU Leuven Iago Cupeiro
KU Leuven Javier Arroyo (now @ LBNL)
SDU Kzrysztof Arendt
LBNL David Blum
NREL Yanfei Li
NREL Kyle Benne
Engie Valentin
LBNL Antoine Gautier
PNNL Sen Huang
UniZG Tea Zakula
Agenda is approved
Assistant professor from Zagreb University, now @ LBNL for 6 months. She has a PhD @ MIT on MPC and has been working on MPC for 10 years. Implementing multi-zone inverse comprenhensive room transfer functions models (miCRTF). These models have similarities with RC grey-box models but their parameters are easier to estimate. Tea and her team have MPC running in real buildings. Particularly, a 20 zones office building with FCU, fully glazed, concrete slabs between floors. In thisi building they observe that the MPC clearly shifts cooling to night time for optimal efficiency.
Breakout sessions Finalized.
To do: Dave: Emulator Peer Review Guide and Checklist to be sent around.
Filip: modifications made based on review, working on documentation.
Kzrysztof: requests to share Filip’s model (not yet overwrite blocks included) to check the documentation (template is being finalized): is currently available in pull request to IDEAS 996.
Roel and Kzrysztof gave feedback.
The emulators are benchmark models (assigned a fixed weather data file etc …), extra feature is made to change the climate and resize the devices (this is optional). Limit the number of benchmark cases! Focus on emulators attached to one climate each, additional functionalities are for a later phase.
The core KPIs concluded from previous IBPSA meeting are listed, please review the ‘Development requirements and guide’ (action all). The KPI calculator module lives within the kpis folder of the BOPTEST repository. It is shown the main functionality of the KPI calculator prototype software for testcase2.
A module for generating data has also been initiated. This module gathers data from the test case (weather, internal gains, energy price profiles, emission factors ...), it processes the data, ensures proper indexing and key values, and it finally brings it within the wrapped.fmu when deploying a test case. This wrapped.fmu is ultimatelly used by BOPTEST to simulate the emulator model. A convention should be agreed in Aachen regarding the columns and format of the final data set which should be, at the same time, aligned with the different agreed scenarios. A predefined format of such data set will enable the forecast module to retrieve the data required by the users for their optimal controllers.
Time integration can be chosen to be done internally in the modelica model or externally in python (both give the same results for the tested case, communication times are kept the same as the time steps in the model).
Waiting for Jesus’ review. To do: Iago send a reminder.
No updates
Ready for internal review (journal submission)
Filip & Gerald (conference + extension for journal submission). Dave is interested in the results.
Time to close the meeting ;-)