Optimizacija legiranja jekla s pomočjo strojnega učenja

ŠTOSIR, ŽIGA (author), Kukar, Matjaž (mentor)

Cilj diplomske naloge je modeliranje doziranja in optimizacija porabe legirnih dodatkov v proizvodnji jekla. V prvem delu predstavimo proces izdelave jekla in nekaj osnov metalurgije, ki so podlaga za razumevanje zastavljenega problema. V nadaljevanju predstavimo orodja in tehnologije, ki smo jih uporabili pri izdelavi naloge. Sledijo opis, način pridobitve, obdelave in vizualizacije podatkov, modeliranje problema in predstavitev rezultatov. Pri delu smo uporabili orodje Orange s programskim jezikom Python in knjižnico Matplotlib. V zadnjem delu predstavimo zaključek ter možnosti prihranka pri porabi legirnih dodatkov in časa, potrebnega za izdelavo šarže, kar predstavlja velik potencial za nadaljnje delo.

• Language: Slovenian
• Keywords: izdelava jekla, strojno učenje, optimizacija, linearno programiranje, Python
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FRI - Faculty of Computer and Information Science ; FMF - Faculty of Mathematics and Physics
• Year: 2019
• PID: 20.500.12556/RUL-106472
• Publication date in RUL: 26.02.2019

Secondary language

• Language: English
• Title: Optimization of steel alloying with machine learning

Abstract

The aim of this thesis is the optimization of alloy additions in steel making. In the first part of this thesis we present some basics of metallurgy and the process of steel making where our problem originates from. Then we present the tools and technologies used in development, followed by a description on how we processed and modelled the data. After that we then describe the results. In conclusion we summarize our work and present possibilities for improvement and further development.

• Keywords: steel making, machine learning, optimization, linear programming, Python