[pt] O processo de formação de pelotas de minério de ferro consiste na preparação das matérias-primas, formação da pelota crua e endurecimento por meio da queima. O produto final deve ser um material poroso que permita a difusão de gases no forno de redução e que, simultaneamente, resista a compressão, característica relevante durante o transporte e no carregamento do forno. No entanto, durante o tratamento térmico e o transporte podem surgir trincas que comprometem a integridade das pelotas. A discriminação de poros e trincas é, portanto, um importante fator para a análise microestrutural e controle de qualidade do material. A microtomografia de raios-x é uma técnica não destrutiva que gera imagens tridimensionais, o que permite uma visualização completa da pelota. No entanto, a metodologia usual de processamento digital de imagens, baseada em extração de atributos de tamanho e forma, apresenta limitações para discriminar poros de trincas. Redes Neurais Deep Learning são uma alternativa poderosa para classificar tipos de objetos em imagens, utilizando como entrada as intensidades dos pixels e atributos automaticamente determinados pela rede. Após treinar um modelo com os padrões correspondente a cada classe, é possível atribuir cada pixel da imagem a uma das classes presentes, permitindo uma segmentação semântica. Nesta dissertação, otimizou-se uma rede Deep Learning com arquitetura U-Net, usando como conjunto de treinamento poucas camadas 2D da imagem 3D original. Aplicando o modelo à pelota utilizada no treinamento foi possível discriminar poros de trincas de forma adequada. A aplicação do modelo a outras pelotas exigiu a incorporação de camadas destas pelotas ao treinamento e otimização de parâmetros do modelo. Os resultados apresentaram classificação adequada, apesar de apresentar dificuldades de criar um modelo geral para discriminação entre poros e trincas em pelotas de minério de ferro. / [en] The iron ore pellet forming process consists of preparing the raw materials, forming the raw pellet and hardening by firing. The end product must be a porous material which allows gas to diffuse in the blast furnace and at the same time resists compression, which is a relevant feature during transport and loading of the furnace. However, during heat treatment and transport cracks may appear that compromise the integrity of the pellets. The discrimination of pores and cracks is therefore an important factor for microstructural analysis and material quality control. X-ray microtomography is a non-destructive technique that generates three-dimensional images, allowing a full view of the pellet. However, the usual methodology of digital image processing, based on extraction of size and shape attributes, has limitations to discriminate crack from pores. Deep Learning Neural Networks are a powerful alternative to classifying object types in images, using as input the pixel intensities and attributes automatically determined by the network. After training a model with the patterns corresponding to each class, it is possible to assign each pixel of the image to one of the classes present, allowing a semantic segmentation. In this dissertation, a Deep Learning network with U-Net architecture was optimized, using as a training set a few 2D layers of the original 3D image. Applying the model to the pellet used in training it was possible to discriminate cracks pores properly. Application of the model to other pellets required the incorporation of layers of these pellets into the training and optimization of model parameters. The results were adequately classified, despite the difficulty of creating a general model for discrimination between pores and cracks in iron ore pellets.

Luossavaara-Kiirunavaara Aktiebolag (LKAB) är en gruvdriftskoncern med järnmalmsbrytning som huvudsaklig verksamhet. En av produkterna tas fram genom att raffinera järnmalm till pellets vilket är en efterfrågad produkt vid ståltillverkning.   Vid sintring av råkulor till pellets är temperaturen väsentlig för att få så hög oxidation som möjligt av magnetit till hematit och för att få rätt fasthet. För att reglera värmen bör temperaturen mätas bland råkulorna eftersom det är kulornas temperatur som är det viktiga.                                   Projektet går ut på att undersöka möjligheterna att mäta temperatur bland råkulor på gratevagnar som åker genom en pelletsugn med ett permanent mätsystem. Här ställs höga krav på mätsystemet på grund av miljö, portabilitet och drifttid. Miljöanalyser gjordes för att ta reda på vilken påfrestning ett mätsystem bör klara av. Förslag om placering av mätutrustning framtogs genom kartläggning av miljön. En värmeundersökning utfördes av vad hög värme har för påverkan på elektroniska komponenter som kan behövas till ett mätsystem. Det gjordes även en undersökning över vilka komponenter som kan klara av den värme ett mätsystem kan bli utsatt för. Utifrån detta kunde ett prototypmätsystem konstrueras för att mäta temperaturen där ett mätsystem kan placeras på en gratevagn samt för att skapa möjlighet för mätningar med termoelement i bädd. Mätsystemet skapades med trådlös kommunikation, möjlighet till loggning av data och med möjlighet att ansluta tre termoelement av typ S. Olika typer av trådlös kommunikation utvärderades för att se vad som kan vara mest lämpat med tanke på miljö, räckvidd, strålning och strömförbrukning. Strömförsörjning undersöktes för att hitta den mest lämpade strömförsörjningen med längst drifttid och högst funktionalitet. En utredning gjordes för att hitta monteringsmöjligheter av ett mätsystem på en gratevagn. Värmeavskärmning och isolering undersöktes för att få ner temperaturen i kapslingen och för elektroniken i ett mätsystem. Några olika temperaturgivare granskades för att hitta en som klarar av att mäta temperaturer upp till 1500°C och som fysiskt kan monteras på en gratevagn. / Luossavaara-Kiirunavaara Aktiebolag (LKAB) is a mining group with iron ore mining as the main business. One of the products is produced by refining iron ore to iron pellets, which is a quality and environmentally clever product  in steel production.                                     When iron pellets are produced by roasting green pellets the temperature is very important. The goal is to get as much oxidation as possible from hematite to magnetite and to get correct mechanical properties. To control the process it’s important to know the temperature of the green pellets specifically and not the surrounding mechanical structure, ceramics lining in the furnace and the influence from radiation.   The goal of the project is to examine possibilities to measure temperature with a permanent measurement system of green pellets which are transported on grate cars trough pellet plant. The measurement system has to meet some requirements of environment, portability and operating time. Analysis of the environment has been made to get specification of what a measurement system has to withstand. Proposition about assembly location has been investigated by analysis of the environment. A thermal analysis has been made of electronic components, especially electronic components that can be used in a measurement system. A analysis has also been made of which electronic components that are suited for a measurement system and can withstand the heat. A development measurement system has been developed from the results of the analysis. The system has been constructed to measure the temperature of what a system can be exposed of on a grate car. This with wireless and data-logging support and future measurements possibilities with three thermocouples of type S.   Wireless data transfer was examined to see what type is most suited for the task with knowledge of environment, transmission range, radiation and power consumption. Power sources was examined to find a power source with longest working time and highest functionality. Heat insulation and reflective materials has been investigated with the goal to get the temperature in the system enclosure as low as possible. Some different types of temperature sensors that can measure temperatures up to 1500°C and which can be mounted on a grate car has been evaluated.