Application of process monitoring based on inferential measurement approach /
Zaidi Salim
Application of process monitoring based on inferential measurement approach / by Zaidi Salim - Kuantan, Pahang : UMP, 2013 - xiii, 44 p. : ill. ; 30 cm. + 1 CD-ROM
Project paper (Bachelor of Chemical Engineering) -- Universiti Malaysia Pahang – 2013
Bibliography : p.40-42
In this study, a new multivariate method to monitor continuous processes is developed based on the Process Control Analysis (PCA) framework. The objective of the study is to develop A new MSPM method and analyze the monitoring performance of system A and B. In industrial practice, monitoring process are usually performed based on an approximate model. As the number of variables increases, the fault detection performance tends to be slow in progression, as well as, introduce greater complexity in the later stages especially in fault identification and diagnosing. These research implements and analyzes Multiple Linear Regression (MLR) method to a continuous process which simplify the number of variables used. This research also based on the conventional MSPM technique. After that, the developed method was analyzed and finally, all the performance result of the developed method was compared with the conventional method. The monitoring results clearly demonstrate the superiority of the proposed method. The MLR methods show that the fault detection performance improved and better than the conventional method.
THE0004995(Local)
Process control--Statistical methods
Regression analysis
Multivariate analysis
Mathematical statistics
Application of process monitoring based on inferential measurement approach / by Zaidi Salim - Kuantan, Pahang : UMP, 2013 - xiii, 44 p. : ill. ; 30 cm. + 1 CD-ROM
Project paper (Bachelor of Chemical Engineering) -- Universiti Malaysia Pahang – 2013
Bibliography : p.40-42
In this study, a new multivariate method to monitor continuous processes is developed based on the Process Control Analysis (PCA) framework. The objective of the study is to develop A new MSPM method and analyze the monitoring performance of system A and B. In industrial practice, monitoring process are usually performed based on an approximate model. As the number of variables increases, the fault detection performance tends to be slow in progression, as well as, introduce greater complexity in the later stages especially in fault identification and diagnosing. These research implements and analyzes Multiple Linear Regression (MLR) method to a continuous process which simplify the number of variables used. This research also based on the conventional MSPM technique. After that, the developed method was analyzed and finally, all the performance result of the developed method was compared with the conventional method. The monitoring results clearly demonstrate the superiority of the proposed method. The MLR methods show that the fault detection performance improved and better than the conventional method.
THE0004995(Local)
Process control--Statistical methods
Regression analysis
Multivariate analysis
Mathematical statistics