Generating Diagnostic Bayesian Networks from Qualitative Causal Models

Denis Kirchhübel; Thomas Martini Jørgensen

Generating Diagnostic Bayesian Networks from Qualitative Causal Models

Denis Kirchhübel, Thomas Martini Jørgensen

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

410 Downloads (Orbit)

Abstract

Safety and efficiency of modern industrial plants can be improved by providing operators with effective digital assistants to diagnose abnormal situations occurring in the plant. To make sense of a large number of alarms, root cause analysis can help pinpoint the origin of an abnormal situation. We investigate the translation of qualitative causal models into Bayesian belief networks (BBN) to utilize efficient tools for probability inference. The diagnosis result of a fault scenario of the Tennessee-Eastman-Process highlight the feasibility of the principle approach and the ongoing research aims to fully leverage the potential of BBN.

Original language	English
Title of host publication	Proceedings of the 2019 24th IEEE International Conference on Emerging Technologies and Factory Automation
Number of pages	6
Publisher	IEEE
Publication date	2019
ISBN (Electronic)	978-1-7281-0303-7
Publication status	Published - 2019
Event	24th IEEE International Conference on Emerging Technologies and Factory Automation - Paraninfo de la Universidad de Zaragoza, Zaragoza, Spain Duration: 10 Sept 2019 → 13 Sept 2019 http://www.etfa2019.org/

Conference

Conference	24th IEEE International Conference on Emerging Technologies and Factory Automation
Location	Paraninfo de la Universidad de Zaragoza
Country/Territory	Spain
City	Zaragoza
Period	10/09/2019 → 13/09/2019
Internet address	http://www.etfa2019.org/

Keywords

Bayesian methods
Expert systems
Process Control
Fault diagnosis
Fault-trees

Access to Document

ETFA paperFinal published version, 1.16 MB

OpenUrl availability

Full text

Cite this

@inproceedings{d99f230500a14241b029bd0deaf9ad0a,

title = "Generating Diagnostic Bayesian Networks from Qualitative Causal Models",

abstract = "Safety and efficiency of modern industrial plants can be improved by providing operators with effective digital assistants to diagnose abnormal situations occurring in the plant. To make sense of a large number of alarms, root cause analysis can help pinpoint the origin of an abnormal situation. We investigate the translation of qualitative causal models into Bayesian belief networks (BBN) to utilize efficient tools for probability inference. The diagnosis result of a fault scenario of the Tennessee-Eastman-Process highlight the feasibility of the principle approach and the ongoing research aims to fully leverage the potential of BBN.",

keywords = "Bayesian methods, Expert systems, Process Control, Fault diagnosis, Fault-trees",

author = "Denis Kirchh{\"u}bel and J{\o}rgensen, \{Thomas Martini\}",

year = "2019",

language = "English",

booktitle = "Proceedings of the 2019 24th IEEE International Conference on Emerging Technologies and Factory Automation",

publisher = "IEEE",

address = "United States",

note = "24th IEEE International Conference on Emerging Technologies and Factory Automation, ETFA 2019 ; Conference date: 10-09-2019 Through 13-09-2019",

url = "http://www.etfa2019.org/",

}

TY - GEN

T1 - Generating Diagnostic Bayesian Networks from Qualitative Causal Models

AU - Kirchhübel, Denis

AU - Jørgensen, Thomas Martini

PY - 2019

Y1 - 2019

N2 - Safety and efficiency of modern industrial plants can be improved by providing operators with effective digital assistants to diagnose abnormal situations occurring in the plant. To make sense of a large number of alarms, root cause analysis can help pinpoint the origin of an abnormal situation. We investigate the translation of qualitative causal models into Bayesian belief networks (BBN) to utilize efficient tools for probability inference. The diagnosis result of a fault scenario of the Tennessee-Eastman-Process highlight the feasibility of the principle approach and the ongoing research aims to fully leverage the potential of BBN.

AB - Safety and efficiency of modern industrial plants can be improved by providing operators with effective digital assistants to diagnose abnormal situations occurring in the plant. To make sense of a large number of alarms, root cause analysis can help pinpoint the origin of an abnormal situation. We investigate the translation of qualitative causal models into Bayesian belief networks (BBN) to utilize efficient tools for probability inference. The diagnosis result of a fault scenario of the Tennessee-Eastman-Process highlight the feasibility of the principle approach and the ongoing research aims to fully leverage the potential of BBN.

KW - Bayesian methods

KW - Expert systems

KW - Process Control

KW - Fault diagnosis

KW - Fault-trees

M3 - Article in proceedings

BT - Proceedings of the 2019 24th IEEE International Conference on Emerging Technologies and Factory Automation

PB - IEEE

T2 - 24th IEEE International Conference on Emerging Technologies and Factory Automation

Y2 - 10 September 2019 through 13 September 2019

ER -

Generating Diagnostic Bayesian Networks from Qualitative Causal Models

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this