Change search
ReferencesLink to record
Permanent link

Direct link
Carthago: development of a tool for the prediction of quality in a glass production furnace
2007 (English)Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

The control of glass manufacturing, as all the industrial processes where many variables are involved, entails high level of complexity. Although this process is completely automated and physically modelled, customers require every day more and more quality. Physical models have reached their limit as it is not possible to measure some of the data on-line with enough accuracy. Data based modelling comes up as method to increase the current accuracy. One of the main problems in glass production is the concentration of bubbles. When high performance glasses are involved, for example for glass-ceramic cookers, a bubble can be the cause of rejection of a complete set. Although it is known that there are some variables capable to cause bubbles, thermodynamics and physics require so many simplifications that they are not directly applicable. The aim of this project is the development of a model for predicting the level of air bubbles in a glass production process by the use of data based methods. The main result of the project is the development of a tool, able to predict the value of five variables using data mining techniques. The results of the predictions are going to be applied to increase the glass quality. As prerequisite of the project, Support Vector Machine algorithm should be used. The tool developed in this project should replace an old application which is currently in use and increase the quality and accuracy of the predictions. As all industrial facilities, a glass furnace is susceptible to having small changes in its functioning as result of time going by. Consequently, the variable metrics could change in the future and decrease the prediction capacity of the tool. Therefore, the tool gives the possibility to retrain the model with new data.

Place, publisher, year, edition, pages
Keyword [en]
Technology, Automatic learning machine, Data mining, Support Vector, Machine, Support Vector Regression
Keyword [sv]
URN: urn:nbn:se:ltu:diva-51748ISRN: LTU-EX--07/152--SELocal ID: 8ee9ef70-929c-4b92-b4e5-4b7816c41a47OAI: diva2:1025112
Subject / course
Student thesis, at least 30 credits
Educational program
Computer Science and Engineering, master's level
Validerat; 20101217 (root)Available from: 2016-10-04 Created: 2016-10-04Bibliographically approved

Open Access in DiVA

fulltext(986 kB)0 downloads
File information
File name FULLTEXT01.pdfFile size 986 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

ReferencesLink to record
Permanent link

Direct link