Stochastic model-predictive control of district-scale building energy systems using SpineOpt

Topi Rasku; Ala Hasan

doi:10.26868/25222708.2023.1262

Stochastic model-predictive control of district-scale building energy systems using SpineOpt

DOES - Design and Operation of Energy Systems

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

64 Downloads (Pure)

Abstract

Model predictive control of buildings is a vibrant field, but mostly focuses on buildings as “pricetakers”. While simplified resistance-capacitance building models have previously been employed within large-scale energy system frameworks to analyse market impacts, tools for stochastic programming are scarce. In this work, we demonstrate the viability of the SpineOpt energy system modelling framework for stochastic model predictive control of an imaginary six-building district using different weather and price forecasts, achieving reasonable performance and cost savings comparable with existing literature. The used methods could be scaled up to city or nation scale energy system studies, or be utilised for electricity market bidding of aggregated building flexibility.

Original language	English
Title of host publication	Proceedings of Building Simulation 2023
Subtitle of host publication	18th Conference of IBPSA
Publisher	International Building Performance Simulation Association (IBPSA)
Pages	893-900
DOIs	https://doi.org/10.26868/25222708.2023.1262
Publication status	Published - 2023
MoE publication type	A4 Article in a conference publication
Event	18th International IBPSA Building Simulation Conference, BS2023 - Radisson Blu Forest Manor Hotel, Shanghai, China Duration: 4 Sept 2023 → 6 Sept 2023 https://bs2023.org/

Publication series

Series	Building Simulation Conference Proceedings
Volume	18
ISSN	2522-2708

Conference

Conference	18th International IBPSA Building Simulation Conference, BS2023
Country/Territory	China
City	Shanghai
Period	4/09/23 → 6/09/23
Internet address	https://bs2023.org/

Funding

This research was funded by the Academy of Finland project Integration of building flexibility into future energy systems (FlexiB) under grant agreement No 332421.

Keywords

model-predictive control
district energy systems
building simulation
resistance-capacitance modelling
Mixed-integer linear programming

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.26868/25222708.2023.1262

FlexiB_BS2023_conference_paper_finalAccepted author manuscript, 3.27 MBLicence: Unspecified

Cite this

@inproceedings{83b3032eadb949019ab518d57d5bc0b3,

title = "Stochastic model-predictive control of district-scale building energy systems using SpineOpt",

abstract = "Model predictive control of buildings is a vibrant field, but mostly focuses on buildings as “pricetakers”. While simplified resistance-capacitance building models have previously been employed within large-scale energy system frameworks to analyse market impacts, tools for stochastic programming are scarce. In this work, we demonstrate the viability of the SpineOpt energy system modelling framework for stochastic model predictive control of an imaginary six-building district using different weather and price forecasts, achieving reasonable performance and cost savings comparable with existing literature. The used methods could be scaled up to city or nation scale energy system studies, or be utilised for electricity market bidding of aggregated building flexibility.",

keywords = "model-predictive control, district energy systems, building simulation, resistance-capacitance modelling, Mixed-integer linear programming",

author = "Topi Rasku and Ala Hasan",

year = "2023",

doi = "10.26868/25222708.2023.1262",

language = "English",

series = "Building Simulation Conference Proceedings",

publisher = "International Building Performance Simulation Association (IBPSA)",

pages = "893--900",

booktitle = "Proceedings of Building Simulation 2023",

address = "Canada",

note = "18th International IBPSA Building Simulation Conference, BS2023 ; Conference date: 04-09-2023 Through 06-09-2023",

url = "https://bs2023.org/",

}

Rasku, T & Hasan, A 2023, Stochastic model-predictive control of district-scale building energy systems using SpineOpt. in Proceedings of Building Simulation 2023 : 18th Conference of IBPSA. International Building Performance Simulation Association (IBPSA), Building Simulation Conference Proceedings, vol. 18, pp. 893-900, 18th International IBPSA Building Simulation Conference, BS2023, Shanghai, China, 4/09/23. https://doi.org/10.26868/25222708.2023.1262

Stochastic model-predictive control of district-scale building energy systems using SpineOpt. / Rasku, Topi; Hasan, Ala.
Proceedings of Building Simulation 2023 : 18th Conference of IBPSA. International Building Performance Simulation Association (IBPSA), 2023. p. 893-900 (Building Simulation Conference Proceedings, Vol. 18).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Stochastic model-predictive control of district-scale building energy systems using SpineOpt

AU - Rasku, Topi

AU - Hasan, Ala

PY - 2023

Y1 - 2023

N2 - Model predictive control of buildings is a vibrant field, but mostly focuses on buildings as “pricetakers”. While simplified resistance-capacitance building models have previously been employed within large-scale energy system frameworks to analyse market impacts, tools for stochastic programming are scarce. In this work, we demonstrate the viability of the SpineOpt energy system modelling framework for stochastic model predictive control of an imaginary six-building district using different weather and price forecasts, achieving reasonable performance and cost savings comparable with existing literature. The used methods could be scaled up to city or nation scale energy system studies, or be utilised for electricity market bidding of aggregated building flexibility.

AB - Model predictive control of buildings is a vibrant field, but mostly focuses on buildings as “pricetakers”. While simplified resistance-capacitance building models have previously been employed within large-scale energy system frameworks to analyse market impacts, tools for stochastic programming are scarce. In this work, we demonstrate the viability of the SpineOpt energy system modelling framework for stochastic model predictive control of an imaginary six-building district using different weather and price forecasts, achieving reasonable performance and cost savings comparable with existing literature. The used methods could be scaled up to city or nation scale energy system studies, or be utilised for electricity market bidding of aggregated building flexibility.

KW - model-predictive control

KW - district energy systems

KW - building simulation

KW - resistance-capacitance modelling

KW - Mixed-integer linear programming

UR - http://www.scopus.com/inward/record.url?scp=85179519876&partnerID=8YFLogxK

U2 - 10.26868/25222708.2023.1262

DO - 10.26868/25222708.2023.1262

M3 - Conference article in proceedings

T3 - Building Simulation Conference Proceedings

SP - 893

EP - 900

BT - Proceedings of Building Simulation 2023

PB - International Building Performance Simulation Association (IBPSA)

T2 - 18th International IBPSA Building Simulation Conference, BS2023

Y2 - 4 September 2023 through 6 September 2023

ER -

Stochastic model-predictive control of district-scale building energy systems using SpineOpt