Measures of Topological Relevance based on the Self-Organizing Map: Applications to Process Monitoring from Spectroscopic Measurements
Abstract
In This Work, the Problem of Real-Time Monitoring of Products' Properties from Spectrophotoscopic Measurements is Presented. Light Absorbance Spectra Are Used as Inputs to Soft Sensors that Estimate Outputs Otherwise Difficult to Measure On-Line. to overcome the Issues Associated to Calibrating the Estimation Models from Very Highdimensional Inputs and a Reduced Number of Observations, We Propose to Select Only a Subset of Relevant Inputs Emerging from the Topological Structure of the Data. the Topologically Preserving Representation is Performed using the Self-Organizing Map (Som) and the Relevance Measured from the U-Matrices. Being based on a Selection of Original Spectral Variables, the Resulting Models Retain the Chemical Interpretability of the Underlying System. Moreover, the Approach is Independent on the Regression Model to Be Embedded in the Soft Sensors. in This Paper, the Utility of the Measures of Topological Relevance (Mtr) over the Som is Discussed on Two Full-Scale Problems from Refining and Pharmaceutical Industry.
Recommended Citation
F. Corona et al., "Measures of Topological Relevance based on the Self-Organizing Map: Applications to Process Monitoring from Spectroscopic Measurements," CEUR Workshop Proceedings, vol. 284, Sun SITE Central Europe, Dec 2007.
Department(s)
Engineering Management and Systems Engineering
International Standard Serial Number (ISSN)
1613-0073
Document Type
Article - Conference proceedings
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2024 Sun SITE Central Europe, All rights reserved.
Publication Date
01 Dec 2007
