Normal view MARC view ISBD view

Implementing PCA-based fault detection system based on selected imported variables for continuous-based process / Siti Nur Liyana Ahamd

By:

Siti Nur Liyana Ahamd

Material type: Text

TextPublication details: Kuantan, Pahang : UMP, 2013Description: xiv, 50 p. : ill. ; 30 cm. + 1 CD-ROMISBN:

THE0002030(Local)

Subject(s):

Dissertation note: Project paper (Bachelor of Chemical Engineering) -- Universiti Malaysia Pahang – 2013 Abstract: Nowadays, the production based on chemical process was rapidly expanding either domestically or internationally. To produce the maximum amount of consistently high quality products as per requested and specified by the customers, the whole process must be considering included fault detection. This is to ensure that product quality is achieved and at the same time to ensure that the quality variables are operated under the normal operation. There were several methods that commonly used to detect the fault in process monitoring such as using SPC or MSPC. However because of the MSPC can operated with multivariable continuous processes with collinearities among process variables, this technique was used widely in industry. In MSPC have a few methods that were proposed to improve the fault detection such as PCA, PARAFAC, multidimensional scaling technique, partial least squares, KPCA, NLPCA, MPCA and others. Here, in this thesis was to proposed new technique which was by implementing PCA-based fault detection system based on selected imported variables for continuous-based process. This technique was selected depends on the highest number of magnitude of correlation of variables using Matlab Software. The result in this thesis was the fault can be detected using only selected important variables in the process.

Tags from this library: No tags from this library for this title. Log in to add tags.

Average rating: 0.0 (0 votes)

Holdings ( 2 )
Title notes ( 3 )

Holdings
Item type	Current library	Call number	Copy number	Status	Date due	Barcode
Final Year Report	UMPLIB GAMBANG	QA278.5 .L59 2013 rs Bc. (Browse shelf(Opens below))	1	Not for loan		0000074795
Final Year Report	UMPLIB GAMBANG	CD 7097 \| QA278.5 .L59 2013 rs Bc. (Browse shelf(Opens below))	1	Not for loan		0000074796

Project paper (Bachelor of Chemical Engineering) -- Universiti Malaysia Pahang – 2013

Bibliography : p. 42-44

Nowadays, the production based on chemical process was rapidly expanding either domestically or internationally. To produce the maximum amount of consistently high quality products as per requested and specified by the customers, the whole process must be considering included fault detection. This is to ensure that product quality is achieved and at the same time to ensure that the quality variables are operated under the normal operation. There were several methods that commonly used to detect the fault in process monitoring such as using SPC or MSPC. However because of the MSPC can operated with multivariable continuous processes with collinearities among process variables, this technique was used widely in industry. In MSPC have a few methods that were proposed to improve the fault detection such as PCA, PARAFAC, multidimensional scaling technique, partial least squares, KPCA, NLPCA, MPCA and others. Here, in this thesis was to proposed new technique which was by implementing PCA-based fault detection system based on selected imported variables for continuous-based process. This technique was selected depends on the highest number of magnitude of correlation of variables using Matlab Software. The result in this thesis was the fault can be detected using only selected important variables in the process.