WP1.2 Coordination meeting January 16, 2018

Meeting Report WP1.2 ‘Modelica library for MPC’

1 MEETING SUBJECT, DATE

Subject: WP1.2

Date: 16-01-2018

Location: Skype for Business

Minutes taken by: Javier Arroyo (KU Leuven)

2 PARTICIPANTS

Company/Organisation Participant

KU Leuven Lieve Helsen (WP Leader)

KU Leuven Iago Cupeiro

KU Leuven Javier Arroyo

KU Leuven Filip Jorissen

SDU Kzrysztof Arendt

LBNL David Blum

ENGIE Lab Valentin Gavan

ENGIE – Axima (now at LBNL) Lisa Rivalin

Pacific Noerthwest National Laboratory Draguna Vrabie

Pacific Noerthwest National Laboratory Sen Huang

SDU Christian Veje

3E (still not decided if will join the project) Roel De Coninck (as observer)

3 AGENDA

Chair: Lieve Helsen

Secretary: Javier Arroyo

Task Leaders:

T1.2.1: Javier Arroyo

T1.2.2: Iago Cupeiro

T1.2.3: Dave Blum/Michael Wetter

Main Aim: preparation Berlin meeting

 Introduction and welcome new partner: Pacific Northwest National Laboratory (Draguna Vrabie & Sen Huang)

 Berlin meeting: draft agenda, see https://github.com/ibpsa/project1/wiki/2018-02-27-expert-meeting-agenda To be prepared w.r.t. WP1.2:

 Plenary: WP 1.2 Model Predictive Control (10 min, Lieve)

 Plenary: What expertize is missing? (all)

 Plenary: Input Project 1 publications (all)

 Day 1 Break-out session (50 min): content & leader

 Day 1 Break-out session (90 min): content & leader

 Day 2 Break-out session (45 min): content & leader

 Day 2 Break-out session (75 min): content & leader

 Summary and action items of break-out groups: WP 1.2 Model Predictive Control (Lieve)

 Information in google docs: decision what to present and discuss at the Berlin meeting (February)

 Available emulator models

https://docs.google.com/document/d/1uT17uEteoXmaDZVYm0DxGODje4Pxfe7oqKPCF4oIhmM/edit?usp=sharing

 Performance indicators for MPC Toolchain and Algorithm evaluation

https://docs.google.com/document/d/1cON-KdJ7BFzSFODDW3JfYeLIm5iWKtTaktCOU1wgRF0/edit#heading=h.ub8zw98vrz07

 MPC formulations

https://docs.google.com/document/d/1dyDA1t4eXMeW7-IvmvAW4D8UDeOEypeAq4CrqyTjdtM/edit?usp=sharing

 Extra optimization experts.

https://docs.google.com/document/d/1twdi04jCLmhYcPEidqQvysYiemmekyYyNOaFc0B698I/edit?usp=sharing

 Other issues?

 Next meeting (WP Leader and Task Leaders to check status break-out sessions)

4 DISCUSSIONS

4.1 INTRODUCTION AND WELCOME NEW PARTNER

Lieve introduces the goal of the skype meeting: to develop a detailed schedule for the Berlin meeting. She welcomes Roel from 3E (Belgium), who is attending as an observer since they are in the negotiation phase to decide whether they will join the project or not. She introduces Sen and Draguna from the Pacific Northwest National Laboratory (USA). They have decided to join the project in this work package. They roughly went through the planning but have already given some thoughts on how they can contribute.

4.2 PREPARATION OF A DRAFT AGENDA FOR BERLIN

• Lieve: We need to organize our input to the plenary sessions and the WP1.2 break-out sessions. In the plenary sessions we will:

 shortly present ourselves,

 shortly present WP2.1 (action Lieve),

 contribute to the discussion about what expertize is missing (action all),

 elaborate on which publications we plan to deliver (action all),

 Summary and action items of WP1.2 break-out group (action Lieve).

The main WP1.2 issues will be discussed during the four break-out sessions. It is important to prepare these in advance such that we can take the most out of it. The draft agenda of the Berlin meeting (https://github.com/ibpsa/project1/wiki/2018-02-27-expert-meeting-agenda) proposes 2 break-out sessions on day 1 (50 minutes and 90 minutes), and 2 break-out sessions on day 2 (45 minutes and 75 minutes). We need to make a draft agenda for these sessions, taking the information available through our google docs as a starting point (emulators, MPC formulations, performance indicators). Lieve wants to attend all four break-out sessions to make a summary/synthesis at the end, so an additional session chair is assigned for each break-out session. Break-out session – 1 & 2 EMULATOR MODELS: 14 lines are completed in the google doc. In total we need 10 emulators: an air-based and hydronic system in each of the following

1. Single zone residential building
2. Multi-zone residential building
3. Single zone office building (was added because of typical buildings in the US)
4. Multi-zone office building
5. Multi-zone office building with hybrid systems

Proposal is that these candidate emulator models are shortly (max 8 minutes per model) presented in the break-out session, followed by a discussion on: (1) selection of the ten emulators for benchmarking, (2) template of documentation, (3) level of model detail needed (allowing to test both high level and low level MPC formulations), (4) roadmap and timeline. Given the number of candidate models, we will need the first two break-out sessions for this (50 + 90 minutes).

• Christian: wouldn’t it be interesting to give everybody the chance of 10 minutes to present themselves?
• Lieve: that is planned for the plenary session.
• Christian: I mean within the work package. Specify the background, interests…
• Lieve: that would take a whole break-out session… I agree this is interesting but I believe it is more important to talk about the WP1.2 tasks and to get started.
• Filip: what about a presentation of 5 minutes for each partner?
• Lieve: OK, a 5 minutes presentation per institute as a whole/team. We will start with that. I count 6 institutes, so around 30 minutes, we will leave 40 minutes for this. Each institute is asked to prepare short presentations of the emulators gathering more information than the already available in the google doc.
• Lisa: We will probably upload a new emulator model that is not yet in the list.
• Krysztof: Regarding the emulators, we have proposed two models that are not ready but that we plan to develop since we have the measurements. So we will present our possibilities there. These models will be in Modelica.
• Lieve: Javier will lead the first session as task leader of T1.2.1. Break-out session - 3 MPC FORMULATION: 5 lines are completed in the google doc. Proposal is that these different MPC formulations are shortly (max 5 minutes per MPC formulation) presented in the break-out session, followed by a discussion. The aim of Task 1.2.2 is not to select the best MPC approach, but rather to get insights and formulate recommendations based on these insights.
• Lieve: Iago will lead the second session as task leader of T1.2.2.
• David: I fully agree, we should give recommendations depending on the application or type of building.
• Iago: we should present our different formulations and highlight which are the strengths and weaknesses of each.
• David: I do have one concern about the MPC formulations. 5 minutes each is too short… it is going to take longer, maybe we should discuss less time the KPI or spread some topics through the sessions… PERFORMANCE INDICATORS (this part can partly shift to session 4): The google doc shows a nice list of performance indicators.
• Lieve: I’m very happy to see the table for KPI, we have a good starting point there. the discussion should rather focus on the mathematical formulation and quantification of these performance indicators, because some of them are difficult to quantify, like the flexibility indicator. Break-out session – 4 LIBRARY for MPC: we already have the Modelica library for simulation models (IBPSA library). We would like to extend this with a library for optimization models (Task 1.2.3), which will be used for MPC in our context. For this we need to define the requirements and develop an approach.
• Lieve: For the MPC library Filip wants to make a first draft since he already has some thoughts and discussions about it.
• Filip:. Whom I could contact to get some preliminary feedback.
• David: You can contact me to already exchange some ideas regarding the organisation and prepare some open-form beforehand. Almost like these google docs and use it as a starting point. PUBLICATION & DISSEMINATION: what are the final goals of our three tasks? What do we want to deliver at the end? Proposal is that (1) the BOP-TEST benchmarks may lead to new ASHRAE standards that can be used by the whole world to test controllers, (2) the recommendations regarding MPC formulations may lead to a booklet that describes a code of good practice, (3) the MPC library will be an open source library.
• Filip: we could also have, if there is interest, a look at didactical material. Optimization is usually a concept very difficult to grasp, hence, we should provide some simple and illustrative examples.
• Iago: communication and dissemination would fit in the 4th break out session leaving the 3rd for the KPI.
• Lieve: Filip will lead the last break-out session.

4.3 STEPS TO BE TAKEN BEFORE THE BERLIN MEETING

Since all session chairs are from KU Leuven, Lieve will organize an internal meeting to finalize the draft agenda for the break-out sessions, and prepare these break-out sessions in advance. This draft agenda will be sent to all WP1.2 partners for approval or comments/suggestions. The final version will be sent to Michael Wetter. In order to allow us to make a first draft, all partners are asked to complete a new google doc (spreadsheet: https://docs.google.com/spreadsheets/d/1XOoNhrldrCOdAGpj-FFNVK8lW9gpcczSnliOjXCcyxk/edit?usp=sharing) with the presentations you want to give during the break-out sessions or discussions you would like to add. The time available for each slot is indicated. Please add a comment if you need more or less (taking the time constraints into account ;-)). The Berlin meeting, and thus the agenda of the break-out sessions, should allow us to focus on the things that assist us all to get started. That is also the reason why external experts will not yet be invited for the Berlin meeting to give an inspiring talk. This can be organized in a later meeting, we think it is better to have our own focus clear first to get started.

4.4 NEXT MEETING

• Lieve: an internal meeting will be organised at KU Leuven with the break-out session chairs to prepare these sessions. Do you feel another WP1.2 skype meeting is needed before the Berlin meeting?
• Filip: I don’t think so. -The rest of the partners also agreed.-
• Lieve: we will send you the minutes of this meeting, together with the link to a google doc to be completed by all of you to allow us to set up the draft agenda for the break-out sessions ASAP. The draft agenda of the break-out sessions will then be composed based on your input. Thanks all for your contribution!