Process monitoring and predictive diagnostics

undergraduate thesis

• Title (croatian): Nadziranje i prediktivna dijagnostika procesa
• Author: Adriana Brzović
• Mentor: Nenad Bolf (mentor)
• Committee member: Nenad Bolf (predsjednik povjerenstva)
• Committee member: Aleksandra Sander (član povjerenstva)
• Committee member: Željka Ujević Andrijić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Chemical Engineering and Technology; Zagreb
• Defense date and country: 2019-09-19, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Chemical Engineering

Abstract

As technology advances, particularly in the IT sector, more and more companies running large-scale industrial operations look to computer based solutions to replace expensive monitoring and maintenance methods. Machine learning is one such solution gaining popularity in the last couple of years. This paper explores how machine learning models can be used for preventative and predictive maintenance of an industrial heat exchanger by continuously detecting fouling formation. The results of the neural network model show successful prediction of fouling formation in the heat exchanger. Random forest model does not successfully predict fouling formation as it cannot extrapolate on new data. The results of this study show great potential in application of machine learning methods in the industry for improving production processes in terms of greater efficiency, energy and raw material saving.

Abstract (croatian)

Uz napredak tehnologije, posebno IT sektora, javlja se sve veća potreba za računalnim rješenjima koja bi zamijenila skupe metode nadziranja i održavanja procesa u velikim proizvodnim pogonima. Jedno od takvih rješenja s rastućom popularnošću u zadnjih nekoliko godina je strojno učenje. Ovaj rad istražuje primjenu modela strojnog učenja za nadziranje i prediktivnu dijagnostiku industrijskog izmjenjivača topline pomoću kontinuiranog praćenja nastajanja naslaga. Razvijen je model neuronske mreže i model slučajnih šuma. Rezultati razvijenog modela neuronske mreže pokazuju uspješno predviđanje nastajanja naslaga u izmjenjivaču topline. Model slučajnih šuma ne predviđa uspješno nastajanje naslaga zbog nemogućnosti ekstrapoliranja na novim podatcima. Rezultati ostvareni u ovom radu pokazuju veliki potencijal primjene metoda strojnog učenja u industriji za poboljšanje proizvodnih procesa u pogledu veće učinkovitosti, uštede energije i sirovina.

• Keywords: machine learning; heat exchanger; fouling; neural network; random forest
• Keywords (croatian): strojno učenje; izmjenjivač topline; naslage; neuronska mreža; model slučajnih šuma
• Language: english
• URN:NBN: urn:nbn:hr:149:648802
• Study programme: Title: Chemical Engineering - Undergraduate study; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka kemijskog inženjerstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka kemijskog inženjerstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-02-14 08:42:35