Title :
Application of Bayesian belief networks to fault detection and diagnosis of industrial processes
Author :
Azhdari, Maryam ; Mehranbod, Nasir
Author_Institution :
Sch. of Chem., Pet. & Gas Eng., Shiraz Univ., Shiraz, Iran
Abstract :
In industrial processes, to confide the success of planed operation, implementing early and accurate method for recognizing abnormal operating conditions, known as faults, is essential. Effective method for fault detection and diagnosis helps reducing impact of these faults, extols the safety of operation, minimizes down time and reduces manufacturing costs. In this paper, application of BBNs is studied for a benchmark chemical industrial process, known as, Tennessee Eastman in order to achieve early fault detection and accurate probable diagnosis of their causes. Application of Bayesian belief networks for fault detection and diagnosis of Tennessee Eastman process in the graphical context description has not been tested yet. Success of this feature confirms capability and ease use of it as a diagnostic system in actual industrial processes.
Keywords :
belief networks; chemical industry; fault diagnosis; Bayesian belief networks; Tennessee Eastman; benchmark chemical industrial process; fault detection; fault diagnosis; Bayesian methods; Chemical engineering; Fault detection; Fault diagnosis; Monitoring; Process control; Testing; Bayesian Belief Networks (BBNs); Fault Detection; Fault Diagnosis; Tennessee Eastman Process;
Conference_Titel :
Chemistry and Chemical Engineering (ICCCE), 2010 International Conference on
Conference_Location :
Kyoto
Print_ISBN :
978-1-4244-7765-4
Electronic_ISBN :
978-1-4244-7766-1
DOI :
10.1109/ICCCENG.2010.5560369
