Nonlinear partial least square (NPLS) methods with generalized likelihood ratio test (GLRT) for fault detection and diagnosis of chemical processes

Chiranjivi Botre, Majdi Mansouri, Mohamed N. Nounou, Hazem N. Nounou, M. Nazmul Karim

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We presented the problem of fault detection using kernel partial least square (PLS) -based generalized likelihood ratio test (GLRT) and neural net partial least square (PLS) -based GLRT. • TEP results demonstrate the effectiveness of the KPLS -based GLRT technique for detection of multiple faults with low false alarm rate and early fault detection • KPLS regression model is used to predict concentration of the product from online process variable.

Original languageEnglish
Title of host publicationFuels and Petrochemicals Division 2016 - Core Programming Area at the 2016 AIChE Spring Meeting and 12th Global Congress on Process Safety
PublisherAIChE
Pages201-215
Number of pages15
ISBN (Electronic)9781510824942
Publication statusPublished - 2016
Externally publishedYes
EventFuels and Petrochemicals Division 2016 - Core Programming Area at the 2016 AIChE Spring Meeting and 12th Global Congress on Process Safety - Houston, United States
Duration: 10 Apr 201614 Apr 2016

Publication series

NameFuels and Petrochemicals Division 2016 - Core Programming Area at the 2016 AIChE Spring Meeting and 12th Global Congress on Process Safety

Conference

ConferenceFuels and Petrochemicals Division 2016 - Core Programming Area at the 2016 AIChE Spring Meeting and 12th Global Congress on Process Safety
Country/TerritoryUnited States
CityHouston
Period10/04/1614/04/16

Fingerprint

Dive into the research topics of 'Nonlinear partial least square (NPLS) methods with generalized likelihood ratio test (GLRT) for fault detection and diagnosis of chemical processes'. Together they form a unique fingerprint.

Cite this