Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Distillation Column Malfunctions Identification Using SVM Classifier Based on Higher Order Statistics


Affiliations
1 Engineering Department, Nuclear Research Center, Egyptian Atomic Energy Authority, Egypt
2 Electronics and Electrical Communications Department, Menoufia University, Menouf, Egypt
     

   Subscribe/Renew Journal


This paper presents a proposed approach for distillation column malfunction identification using Higher Order Statistics (HOS). Gamma ray scanning techniques have been used for examining the inner details of a distillation column. In the proposed method, the signals are firstly divided into frames; each frame contains only the signal of one column tray. Secondly, HOS are estimated for these frame signals. Thirdly, features are extracted from the HOS estimate. Finally, features are used for training and testing of Support Vector Machine classifier to identify the distillation column malfunctions. The simulation results show that the HOS can be used efficiently for the distillation column malfunction identification especially at high noisy scanning conditions.

Keywords

Bispectrum, Cumulant, Moment, and Trispectrum.
User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 263

PDF Views: 2




  • Distillation Column Malfunctions Identification Using SVM Classifier Based on Higher Order Statistics

Abstract Views: 263  |  PDF Views: 2

Authors

M. E. Hammad
Engineering Department, Nuclear Research Center, Egyptian Atomic Energy Authority, Egypt
H. Kasban
Engineering Department, Nuclear Research Center, Egyptian Atomic Energy Authority, Egypt
O. Zahran
Electronics and Electrical Communications Department, Menoufia University, Menouf, Egypt
M. I. Dessouky
Electronics and Electrical Communications Department, Menoufia University, Menouf, Egypt
S. M. Elaraby
Engineering Department, Nuclear Research Center, Egyptian Atomic Energy Authority, Egypt

Abstract


This paper presents a proposed approach for distillation column malfunction identification using Higher Order Statistics (HOS). Gamma ray scanning techniques have been used for examining the inner details of a distillation column. In the proposed method, the signals are firstly divided into frames; each frame contains only the signal of one column tray. Secondly, HOS are estimated for these frame signals. Thirdly, features are extracted from the HOS estimate. Finally, features are used for training and testing of Support Vector Machine classifier to identify the distillation column malfunctions. The simulation results show that the HOS can be used efficiently for the distillation column malfunction identification especially at high noisy scanning conditions.

Keywords


Bispectrum, Cumulant, Moment, and Trispectrum.