Auxiliary codes for fault prognosis of Tennessee Eastman process using a hybrid model (CPL1.0)

Research output: Contribution to journalArticleScientificpeer-review

Abstract

CPL1.0 is a Matlab code which can generate fault predictions of Tennessee Eastman (TE) process, based on the open-source toolbox developed by Kevin Murphy in 2005. It facilitates the calculation of Prior Probabilities (PP), Conditional Probabilities (CP), and Likelihood Evidence (LE). These are essential features required for fault prognosis purpose using Hidden Markov Model (HMM) and Bayesian Network (BN) hybrid model. Determination of the CP, PP, and LE is the most time-consuming component in the aforementioned process. The proposed code has the potential to drastically reduce the repetitive computation time thus enabling the researcher to focus on the main goal-oriented outcome. CPL1.0 is implemented as a facilitator to communicate between BN and the HMM in a hybrid fault prediction and prognosis system. The hybrid system can predict ten out of ten selected faults and can accurately prognose eight out of the ten faults.
Original languageEnglish
Pages (from-to)100309
JournalSoftwareX
Volume10
DOIs
Publication statusPublished - 1 Jul 2019
MoE publication typeA1 Journal article-refereed

Keywords

  • Conditional probabilities
  • Prior probabilities
  • Likelihood evidence
  • Bayesian networks

Fingerprint

Dive into the research topics of 'Auxiliary codes for fault prognosis of Tennessee Eastman process using a hybrid model (CPL1.0)'. Together they form a unique fingerprint.

Cite this