Volume 32, Issue 4 (IJIEPR 2021)                   IJIEPR 2021, 32(4): 1-10 | Back to browse issues page

XML Print

1- National Institute of Technology Patna, Patna, Bihar, India
2- National Institute of Technology Patna, Patna, Bihar, India , das.anupam13@gmail.com
Abstract:   (1945 Views)
The article highlights the development of a Non-Gaussian Process Monitoring Strategy for a Steel Billet Manufacturing Unit (SBMU). The non-Gaussian monitoring strategy being proposed is based on Modified Independent Component Analysis (ICA) which is a variant of the widely employed conventional ICA. The Independent Components(IC) being extracted by modified ICA technique are ordered as per the variance explained akin to that of Principle Component Analysis (PCA). Whereas in conventional ICA the variance explained by the ICs are not known and thereby causes hindrance in the selection of influential ICs for eventual building of the nominal model for the ensuing monitoring strategy. Hotelling T2 control chart based on modified ICA scores was used for detection of fault(s) whose control limit was estimated via Bootstrap procedure owing to the non-Gaussian distribution of the underlying data. The Diagnosis of the Detected Fault(s) was carried out by employment of Fault Diagnostic Statistic. The Diagnosis of the Fault(s) involved determination of the contribution of the responsible Process and Feedstock characteristics. The non-Gaussian strategy thus devised was able to correctly detect and satisfactory diagnose the detected fault(s)
Full-Text [PDF 596 kb]   (915 Downloads)    
Type of Study: Research | Subject: Statistical Process Control Statistical Process Control or Quality Control
Received: 2021/10/4 | Accepted: 2021/11/17 | Published: 2021/12/7

Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.