แจ้งเอกสารไม่ครบถ้วน, ไม่ตรงกับชื่อเรื่อง หรือมีข้อผิดพลาดเกี่ยวกับเอกสาร ติดต่อที่นี่ ==>
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
Siriwatida Srirabai. Development of support vector machine based fault detection and diagnosis application to vinyl chloride monomer process. Master's Degree(Chemical Engineering). Kasetsart University. Office of the University Library. : Kasetsart University, 2021.
Development of support vector machine based fault detection and diagnosis application to vinyl chloride monomer process
Name: Siriwatida Srirabai
Classification :.LCCS: TP1180
LCSH:
Vinyl chloride
LCSH:
Ethylene dichloride
Abstract:
Monitoring the process status and identifying the presence of process operational faults are essential for improving the process safety in petrochemical plants that interactions between various process streams and units are associated. However, due to various circumstances, identifying abnormal or fault conditions is difficult in practice. Petrochemical processes are typically on a large scale with complex interconnecting between unit operations and process streams leading to a difficult task for monitoring the process status and identifying the root cause when the process faults occurred. Furthermore, some process variables also have a difference in measured sampling rate,an obstacle to analyses the process for fault detection and detection in real-time. To improve the process safety in petrochemical plants, this paper presents the development of support vector machine (SVM) techniques for realtime detecting and identifying the operational fault cases with an application to the constant rate and multirate signal process. The domain knowledge-based feature selection and the linear SVM hyperparameter one-versus-all and one-versus-one multiclass concept through k-fold cross-validation was investigated in the case constant rate signal process. The moving horizon-based approach and the zeroorder and first-order hold data reconstruction are applied for preparing the uniform input samples. The Pearson correlation is applied for a feature selection in the proposed multirate SVM-FDD framework. The classifier is trained through k-fold cross-validation. The SVM kernel types, kernel hyperparameter parameters, oneversus-all and one-versus-one multiclass concept, and the horizon window size are optimized and investigated. A vinyl chloride monomer plant of ethylene dichloride (EDC) pyrolysis and vinyl chloride monomer (VCM) purification sections with twelve operational fault types are applied as a case study. The software-in-the-loop simulation between MATLAB and UniSim Design dynamic simulator is deployed for evaluating the real-time performance of the proposed multirate SVM fault detection and identification framework.
Kasetsart University. Office of the University Library
Address:
Bangkok
Email:
tdckulib@ku.ac.th
Name: Chanin Panjapornpon
Role:
Thesis Advisor
Created:
2021
Modified:
2025-07-23
Issued:
2025-07-23
วิทยานิพนธ์/Thesis
application/pdf
URL: https://www.lib.ku.ac.th/KUthesis/2564/siriwatida-sri-all.pdf
CallNumber:
TP1180.V48 .S57
eng
DegreeName: Master of Engineering
Level: Master's Degree
Descipline: Chemical Engineering
Grantor: Kasetsart University
©copyrights Kasetsart University
RightsAccess:
ใช้เวลา
0.032331 วินาที
Siriwatida Srirabai
| Title | Contributor | Type |
|---|---|---|
| Development of support vector machine based fault detection and diagnosis application to vinyl chloride monomer process
มหาวิทยาลัยเกษตรศาสตร์ Siriwatida Srirabai | Chanin Panjapornpon | วิทยานิพนธ์/Thesis |
Chanin Panjapornpon