Optimal control of chiller plants using Bayesian network

Ana Carolina Laurini Malara, Sen Huang, Wangda Zuo, Michael D. Sohn, Nurcin Celik

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

Linear regression models trained by the dataset, which contain the optimal values for the control parameters under different operating conditions, have been heavily studied in the literature of chiller plants operation optimization due to their performances with higher speeds. However, the linear regression models face difficulties when a nonlinear input/output relationship is considered. Addressing this challenge, we proposed a Bayesian Network (BN) model (a datadriven and probabilistic graphical model), for the operation optimization of chiller plants. Here, we first introduced the construction of the BN model, and demonstrated its validity on the model predictive control of a condenser water set point for a watercooled chiller plant. Then, we evaluated the performance of the proposed BN model under imperfect prediction of weather conditions and building cooling loads with inputs embedding manually generated errors. The baseline performance was provided through a model-based optimization (MBO) using an exhaustive search. The results show that the proposed BN model could provide energy savings compatible with the one given by the MBO using an exhaustive search for both inputs with and without errors.

Original languageEnglish (US)
Title of host publication14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings
PublisherInternational Building Performance Simulation Association
Pages449-455
Number of pages7
StatePublished - 2015
Event14th Conference of International Building Performance Simulation Association, BS 2015 - Hyderabad, India
Duration: Dec 7 2015Dec 9 2015

Other

Other14th Conference of International Building Performance Simulation Association, BS 2015
CountryIndia
CityHyderabad
Period12/7/1512/9/15

Fingerprint

Bayesian networks
Bayesian Model
Bayesian Networks
Network Model
Optimal Control
Optimization
Exhaustive Search
Linear Regression Model
Model-based
Model Predictive Control
Graphical Models
Energy Saving
Probabilistic Model
Weather
Imperfect
Point Sets
Control Parameter
Cooling
Linear regression
Baseline

ASJC Scopus subject areas

  • Computer Science Applications
  • Architecture
  • Modeling and Simulation
  • Building and Construction

Cite this

Malara, A. C. L., Huang, S., Zuo, W., Sohn, M. D., & Celik, N. (2015). Optimal control of chiller plants using Bayesian network. In 14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings (pp. 449-455). International Building Performance Simulation Association.

Optimal control of chiller plants using Bayesian network. / Malara, Ana Carolina Laurini; Huang, Sen; Zuo, Wangda; Sohn, Michael D.; Celik, Nurcin.

14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings. International Building Performance Simulation Association, 2015. p. 449-455.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Malara, ACL, Huang, S, Zuo, W, Sohn, MD & Celik, N 2015, Optimal control of chiller plants using Bayesian network. in 14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings. International Building Performance Simulation Association, pp. 449-455, 14th Conference of International Building Performance Simulation Association, BS 2015, Hyderabad, India, 12/7/15.
Malara ACL, Huang S, Zuo W, Sohn MD, Celik N. Optimal control of chiller plants using Bayesian network. In 14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings. International Building Performance Simulation Association. 2015. p. 449-455
Malara, Ana Carolina Laurini ; Huang, Sen ; Zuo, Wangda ; Sohn, Michael D. ; Celik, Nurcin. / Optimal control of chiller plants using Bayesian network. 14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings. International Building Performance Simulation Association, 2015. pp. 449-455
@inproceedings{233105c1971e441ca54416687993cf6f,
title = "Optimal control of chiller plants using Bayesian network",
abstract = "Linear regression models trained by the dataset, which contain the optimal values for the control parameters under different operating conditions, have been heavily studied in the literature of chiller plants operation optimization due to their performances with higher speeds. However, the linear regression models face difficulties when a nonlinear input/output relationship is considered. Addressing this challenge, we proposed a Bayesian Network (BN) model (a datadriven and probabilistic graphical model), for the operation optimization of chiller plants. Here, we first introduced the construction of the BN model, and demonstrated its validity on the model predictive control of a condenser water set point for a watercooled chiller plant. Then, we evaluated the performance of the proposed BN model under imperfect prediction of weather conditions and building cooling loads with inputs embedding manually generated errors. The baseline performance was provided through a model-based optimization (MBO) using an exhaustive search. The results show that the proposed BN model could provide energy savings compatible with the one given by the MBO using an exhaustive search for both inputs with and without errors.",
author = "Malara, {Ana Carolina Laurini} and Sen Huang and Wangda Zuo and Sohn, {Michael D.} and Nurcin Celik",
year = "2015",
language = "English (US)",
pages = "449--455",
booktitle = "14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings",
publisher = "International Building Performance Simulation Association",

}

TY - GEN

T1 - Optimal control of chiller plants using Bayesian network

AU - Malara, Ana Carolina Laurini

AU - Huang, Sen

AU - Zuo, Wangda

AU - Sohn, Michael D.

AU - Celik, Nurcin

PY - 2015

Y1 - 2015

N2 - Linear regression models trained by the dataset, which contain the optimal values for the control parameters under different operating conditions, have been heavily studied in the literature of chiller plants operation optimization due to their performances with higher speeds. However, the linear regression models face difficulties when a nonlinear input/output relationship is considered. Addressing this challenge, we proposed a Bayesian Network (BN) model (a datadriven and probabilistic graphical model), for the operation optimization of chiller plants. Here, we first introduced the construction of the BN model, and demonstrated its validity on the model predictive control of a condenser water set point for a watercooled chiller plant. Then, we evaluated the performance of the proposed BN model under imperfect prediction of weather conditions and building cooling loads with inputs embedding manually generated errors. The baseline performance was provided through a model-based optimization (MBO) using an exhaustive search. The results show that the proposed BN model could provide energy savings compatible with the one given by the MBO using an exhaustive search for both inputs with and without errors.

AB - Linear regression models trained by the dataset, which contain the optimal values for the control parameters under different operating conditions, have been heavily studied in the literature of chiller plants operation optimization due to their performances with higher speeds. However, the linear regression models face difficulties when a nonlinear input/output relationship is considered. Addressing this challenge, we proposed a Bayesian Network (BN) model (a datadriven and probabilistic graphical model), for the operation optimization of chiller plants. Here, we first introduced the construction of the BN model, and demonstrated its validity on the model predictive control of a condenser water set point for a watercooled chiller plant. Then, we evaluated the performance of the proposed BN model under imperfect prediction of weather conditions and building cooling loads with inputs embedding manually generated errors. The baseline performance was provided through a model-based optimization (MBO) using an exhaustive search. The results show that the proposed BN model could provide energy savings compatible with the one given by the MBO using an exhaustive search for both inputs with and without errors.

UR - http://www.scopus.com/inward/record.url?scp=84976439460&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84976439460&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84976439460

SP - 449

EP - 455

BT - 14th International Conference of IBPSA - Building Simulation 2015, BS 2015, Conference Proceedings

PB - International Building Performance Simulation Association

ER -